- Email: [email protected]
Effects of nongenetic factors on immune cell dynamics in early childhood: The Generation R Study
Accepted Manuscript Effects of nongenetic factors on immune cell dynamics in early childhood: The Generation R Study Diana van den Heuvel, PhD, Michelle A.E. Jansen, MD, Kazem Nasserinejad, MSc, Willem A. Dik, PhD, Ellen G. van Lochem, PhD, Liesbeth E. Bakker-Jonges, PhD, Halima Bouallouch-Charif, BSc, Vincent W.V. Jaddoe, MD, PhD, Herbert Hooijkaas, PhD, Jacques J.M. van Dongen, MD, PhD, Henriëtte A. Moll, MD, PhD, Menno C. van Zelm, PhD PII:
S0091-6749(16)31379-3
DOI:
10.1016/j.jaci.2016.10.023
Reference:
YMAI 12477
To appear in:
Journal of Allergy and Clinical Immunology
Received Date: 17 April 2016 Revised Date:
29 August 2016
Accepted Date: 5 October 2016
Please cite this article as: van den Heuvel D, Jansen MAE, Nasserinejad K, Dik WA, van Lochem EG, Bakker-Jonges LE, Bouallouch-Charif H, Jaddoe VWV, Hooijkaas H, van Dongen JJM, Moll HA, van Zelm MC, Effects of nongenetic factors on immune cell dynamics in early childhood: The Generation R Study, Journal of Allergy and Clinical Immunology (2016), doi: 10.1016/j.jaci.2016.10.023. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
TE
D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Immunophenotyping of Children Aged 0 - 6 Years Lymphocytes NK cells
Granulo
T cells
AC C
Mono
Lympho
Z- score
CD3
Naive
Naive
Mem
CD8+ T cells
Mem
CD4+ T cells
Mem
B cells Naive
NK cells
Monocytes
Granulocytes
Effects of Nongenetic Determinants on Leukocyte Kinetics 62 Leukocyte Populations
26 Nongenetic Determinants
Birth Characteristics (n=6) Bacterial/Viral Exposure (n=11) Distinct effects of various determinants in the indicated category
Z- score
Maternal Immune-mediated Diseases (n=3)
Decrease, or
Innate Leukocytes (n=8)
2
Naive Lymphocytes (n=24)
3
Memory Lymphocytes (n=18)
2.00
4
Memory Lymphocytes (n=12)
1.00
Maternal Life Style (n=6)
Increase
1
Z- score
CD45
7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00 -1.00 -2.00 3.00 2.50 2.00 1.50 1.00 0.50 0.00 -0.50 -1.00 -1.50 -2.00 3.00 2.00 1.00 0.00 -1.00 -2.00 -3.00 -4.00 -5.00
Z- score
CD16/CD56
EP
All Events
SSC
Kinetics Profiles of 62 Leukocyte Subsets
0.00 -1.00 -2.00 -3.00 -4.00 -5.00
0
6
12 18 24 30 36 42 48 54 60 66 72 Age (months)
ACCEPTED MANUSCRIPT Effects of nongenetic factors on immune cell dynamics in early childhood: The
Running title: Determinants of leukocyte subset kinetics in children
RI PT
Generation R Study
Diana van den Heuvel, PhD,a Michelle A.E. Jansen, MD,b,c Kazem Nasserinejad, MSc,d Willem A. Dik, PhD,a Ellen G. van Lochem, PhD,a# Liesbeth E. Bakker-Jonges, PhD,a‡
M AN US C
Halima Bouallouch-Charif, BSc,a Vincent W.V. Jaddoe, MD, PhD,b,c,e Herbert Hooijkaas, PhD,a Jacques J.M. van Dongen, MD, PhD,a* Henriëtte A. Moll, MD, PhD,c* Menno C. van Zelm, PhDa,f* a
Department of Immunology, Erasmus MC, University Medical Center, Rotterdam, the
Netherlands;
b
The Generation R Study Group, Erasmus MC, University Medical Center,
Rotterdam, the Netherlands; the Netherlands;
d
c
Department of Pediatrics, Erasmus MC-Sophia, Rotterdam,
Department of Biostatistics, Erasmus MC, University Medical Center, e
Department of Epidemiology, Erasmus MC, University
D
Rotterdam, the Netherlands;
Medical Center, Rotterdam, the Netherlands;
f
Department of Immunology and Pathology,
TE
Central Clinical School, Monash University, Melbourne, Victoria, Australia;
#
Present
address: Department of Microbiology and Immunology, Rijnstate Hospital, Arnhem, the
EP
Netherlands; ‡ Present address: Department of Medical Laboratory, Reinier de Graaf Groep, Delft, The Netherlands;
AC C
* shared senior authorship
Correspondence:
Menno C. van Zelm, PhD
Department of Immunology and Pathology, Central Clinical School, Monash University, 89 Commercial Road, Melbourne, Victoria 3004, Australia Phone: +61 39903 0834 Email: [email protected]
This work was supported by an Erasmus MC Fellowship to M.C.v.Z
1
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
anti-TPO, anti-thyroid peroxidase IgG CI, confidence interval class, classical monocytes CMV, cytomegalovirus EBV, Epstein Barr virus eos, eosinophils H pylori, Helicobacter pylori HSV-1, herpes simplex virus 1 inter, intermediate monocytes mem, memory n-class, non-classical monocytes neu, neutrophils PROMM, premature rupture of membranes RTI, respiratory tract infection Tcm, central memory T cells Tem, effector memory T cells TG2A, anti-tissue transglutaminase IgA antibody VZV, varicella zoster virus
RI PT
Abbreviations:
2
ACCEPTED MANUSCRIPT ABSTRACT
2
Background. Numbers of blood leukocyte subsets are highly dynamic in childhood and
3
differ greatly between individuals. Inter-individual variation is only partly accounted for by
4
genetic factors.
5
Objective. Determine which nongenetic factors affect the dynamics of innate leukocytes, and
6
naive and memory lymphocyte subsets.
7
Methods. We performed six-color flow cytometry and linear mixed effect modeling to define
8
the dynamics of 62 leukocyte subsets from birth to 6 years of age in 1,182 children with one
9
to five measurements per individual. Subsequently, we defined the impact of prenatal
10
maternal lifestyle-related or immune-mediated determinants, birth characteristics and
11
bacterial/viral exposure-related determinants on leukocyte subset dynamics.
12
Results. Functionally similar leukocyte populations were grouped by unbiased hierarchical
13
clustering of patterns of age-related leukocyte dynamics. Innate leukocyte numbers were high
14
at birth and were predominantly affected by maternal low education level. Naive
15
lymphocytes peaked around 1 year, while most memory lymphocyte subsets more gradually
16
increased during the first 4 years of life. Dynamics of CD4+ T cells were predominantly
17
associated with gender, birth characteristics, and persistent infections with cytomegalovirus
18
(CMV) or Epstein Barr virus (EBV). CD8+ T cells were predominantly associated with CMV
19
and EBV infections, and TCRγδ+ T cells with premature rupture of membranes and CMV
20
infection. B-cell subsets were predominantly associated with gender, breastfeeding and
21
Helicobacter pylori carriership.
22
Conclusions. Our study identifies specific dynamic patterns of leukocyte subset numbers, as
23
well as nongenetic determinants that affect these patterns, thereby providing new insights into
24
the shaping of the childhood immune system.
AC C
EP
TE
D
M AN US C
RI PT
1
25
3
ACCEPTED MANUSCRIPT 26 27
KEY MESSAGES •
Absolute numbers of leukocyte populations are differ between individuals, but functionally related leukocyte populations follow similar age-related leukocyte dynamics
29
between birth and 6 years of age.
30
•
RI PT
28
We identified nongenetic factors that are associated with the dynamics of innate leukocyte, B-cell and CD4+ and CD8+ T-cell lineages (as well as specific leukocyte
32
subsets) and that underlie, at least in part, the human immunological diversity.
M AN US C
31
33
CAPSULE SUMMARY
35
Human inter-individual immunological diversity can only partly be explained by genetic
36
factors, indicating an important contribution of nongenetic factors. We here analyzed
37
leukocyte dynamics in 1,182 children of the Generation R Study and determined the effect of
38
26 nongenetic factors on inter-individual immunological diversity.
TE
D
34
39
KEY WORDS
41
Longitudinal
42
leukocytes, CD4+ T cells, CD8+ T cells, B cells, TCRγδ+ T cells, nongenetic determinants.
leukocyte
dynamics,
inter-individual
AC C
43
EP
40
4
immunological
variance,
innate
ACCEPTED MANUSCRIPT INTRODUCTION
45
The human immune system shows high diversity in cellular composition and functional
46
responses between individuals. Blood leukocytes in young children are highly dynamic in
47
numbers and composition.1-6 Innate cell numbers, such as neutrophils and NK cells, are
48
higher in neonates than in children or adults,1, 5, 7 and already display dynamic changes within
49
the first few days after birth.1, 5, 6 B and T cells are mostly naive in infants, while protective
50
immunity is gradually built up in the form of increasing numbers of memory B and T cells
51
during the first 5 years of life after which these numbers stabilize.2-4, 8-11 Recent studies have
52
found that 25-50% of inter-individual variation in cellular composition and functional
53
responses were accounted for by genetic factors,12, 13 indicating that nongenetic determinants
54
underlie a large part of immune trait variance,14 which will potentially have long-term effects
55
and underlie part of the immune trait variation in adults.
D
M AN US C
RI PT
44
To date, correlations of patterns between different immune cells are incompletely
57
studied, due to either the restricted numbers of analyzed subsets, the short-term follow-up, or
58
the lack of longitudinal analyses correcting for intrapersonal correlations between repeated
59
measures. Furthermore, the nongenetic determinants that drive the kinetics of each type of
60
immune cell remain less well studied. Likely, these involve various environmental
61
determinants, such as maternal life style, maternal immune-mediated diseases, birth
62
characteristics, and bacterial and viral exposure in childhood.15-24 Studying the impact of
63
these nongenetic factors on the dynamics of childhood immune development requires large
64
cohorts of healthy young children, with multiple measurements per individual.
AC C
EP
TE
56
65
We investigated which nongenetic factors, related to the prenatal maternal lifestyle,
66
prenatal maternal immune-mediated diseases, birth characteristics or bacterial/viral exposure-
67
related characteristics, influence the dynamics of blood leukocyte populations from birth until
5
ACCEPTED MANUSCRIPT 68
6 years of age. This concerned a total of 62 leukocyte populations, including innate leukocyte
69
subsets, naive and memory B-cell and T-cell subsets.
AC C
EP
TE
D
M AN US C
RI PT
70
6
ACCEPTED MANUSCRIPT METHODS
72
Study subjects
73
This study was embedded in the Generation R Study, a prospective population-based cohort
74
study from fetal life until young adulthood.25, 26 The current study focused on a subgroup of
75
1,182 two-generation Dutch children, born between August 2003 and August 2006.
76
Peripheral blood was obtained at birth, and median age of 6 months, 14 months, 25 months
77
and 72 months. Detailed immunophenotyping was performed at 1-5 time points per child,
78
resulting in a total of 2,010 data points (details in Suppl Methods). Written informed consent
79
was obtained from parents, according to the Medical Ethics Committee guidelines of
80
Erasmus MC.
M AN US C
RI PT
71
81
Immunophenotyping
83
Absolute numbers of leukocytes, NK cells, T cells and B cells were obtained with a
84
diagnostic lyse-no-wash protocol, using a BD FACSCalibur. Six-color flow cytometry was
85
performed to quantify 62 well-defined leukocyte populations (Table 1), which included NK
86
cell, monocyte, granulocyte, naive and memory B-cell, naive and memory TCRαβ+ T-cell
87
and TCRγδ+ T-cell subsets (Supplemental Table 1). Flowcytometric analyses were performed
88
on an LSRII (BD Biosciences) using standardized measurement settings (details in Suppl
89
Methods).27
AC C
EP
TE
D
82
90 91
Determinants
92
Information on 22 dichotomized and 4 continuous determinants was obtained (Table 2).25, 26
93
Six determinants were related to prenatal maternal life style, and were evaluated using
94
questionnaires in the first, second and third trimester of pregnancy:26 maternal age and body
95
mass index before pregnancy, education, net household income, and smoking or alcohol use
7
ACCEPTED MANUSCRIPT during pregnancy. Three determinants were related to prenatal maternal immune-mediated
97
diseases: data on maternal atopy was obtained by maternal reported questionnaires during
98
pregnancy,26 serum anti-thyroid peroxidase IgG (anti-TPO) and anti-tissue transglutaminase
99
IgA antibody (TG2A) levels were measured in the second trimester of pregnancy (mean ±
100
SD; 13.4 ± 2.0 weeks; and 20.6 ± 1.2 weeks, respectively).28, 29 Information on 6 birth-related
101
determinants, i.e. gender, gestational age (preterm birth <37w), birth weight (low birth weight
102
<2,500g), premature ruptures of membranes, mode of delivery and birth season was obtained
103
by obstetric records assessed in midwife practices and hospitals.25, 26 Eleven bacterial or viral
104
infection-related determinants were included: breastfeeding and breastfeeding duration,
105
having siblings, antibiotics usage and presence of upper or lower respiratory tract infection in
106
the first year of life (obtained using questionnaires at the child’s age of 2, 3, 6 and 12
107
months),26 Helicobacter pylori (H. pylori) carriership at 6y of age, and seropositivity (IgG)
108
for cytomegalovirus (CMV), Epstein Barr virus (EBV), herpes simplex virus 1 (HSV-1) and
109
varicella zoster virus (VZV) at the age of 6 years.22, 30, 31
110
EP
TE
D
M AN US C
RI PT
96
Statistical modeling
112
To model leukocyte dynamics between birth and the age of 6 years, linear mixed effect
113
analyses were performed. By including random-effects in the model, this approach enabled
114
modeling of cross-sectional data, with further improvement of the accuracy by incorporating
115
longitudinal follow-up data from individual children. To capture the trend in the data more
116
precisely, the age of the children was included as a natural spline with different knots (0-3
117
knots) into the models. Basically, the number of knots is inversely related to the smoothness
118
of the curve. Positions of the knots were defined as the 50th percentile (25.5 months) in the 1-
119
knot model, and the 33rd and 66th percentiles (14.1 and 70 months) in the 2-knots model. The
AC C
111
8
ACCEPTED MANUSCRIPT 120
knots in the 3-knot model were defined manually at 6, 14 and 24 months, focusing around the
121
time points of data measurement. Model selection was performed by likelihood ratio test. Next, for each leukocyte population the effect of all 26 determinants on the models was
123
assessed by first adding each determinant univariably into the model and analyzing fixed
124
effect estimates. To correct for potential multiple testing errors, a correction for the four
125
groups of determinants was performed, and consequently only determinants with an effect of
126
p<0.0125 were defined as significant. Subsequently, for each leukocyte population, the
127
determinants with a significant effect (up to 3 determinants were significant per model) were
128
combined in a multivariable model to correct for possible confounding effects. Last, to test
129
whether changes in leukocyte dynamics over time (from birth until 6 years), induced by the
130
determinant of interest was most noticeable at a specific age interval the individual
131
determinants were tested in relation to the age of the children in a multivariable model. The
132
p-value for significance for the multivariable analyses was p<0.05. Statistical analyses were
133
performed in R (version R-3.2.1; details in Supplemental Material and Methods).32
134
EP
TE
D
M AN US C
RI PT
122
Clustering analyses
136
To cluster patterns of leukocyte subsets kinetics, the modelled data of each population was
137
normalized into zero mean and unit standard deviation (z-score), using the following
138
calculation:
139
z-score =
140
The normalized leukocyte models were subsequently clustered using agglomerative (‘bottom-
141
up’) Ward’s hierarchical clustering, at each step clustering two clusters with minimum
142
between-cluster distance, using the Euclidean distance measure.
AC C
135
9
ACCEPTED MANUSCRIPT RESULTS
144
Linear mixed effect modeling of leukocyte-subset cell numbers versus the child’s age
145
To study the dynamics of blood immune cells in young children, we quantified cell numbers
146
of 11 leukocyte subsets in 1,182 children between birth and 76 months (6y) of age. Statistical
147
modeling showed that the numbers of four innate leukocyte subsets, i.e. monocytes,
148
neutrophils, eosinophils and NK cells, were high at birth, quickly declined within the first 6
149
months of age and subsequently remained stable (Figure 1A). Naive B-cell and T-cell
150
numbers strongly increased after birth, peaked between 6-14 months of age and subsequently
151
decreased to stable levels around the age of 2-6 years (Figure 1B). Memory B-cell and T-cell
152
numbers slowly increased within the first 6-14 months of life, after which numbers declined
153
marginally and stabilized from ~3 years onwards (Figure 1B). Total TCRγδ+ T-cell numbers
154
increased until 6 months and remained quite stable at these levels (Figure 1C).
D
M AN US C
RI PT
143
TE
155
Distinct dynamics of innate leukocyte, and naive and memory B and T cells
157
To evaluate whether functionally related immune populations would follow similar dynamics
158
with increasing age, we extended our analysis in the same children to 62 leukocyte subsets,
159
based on recent insights into naive and memory B-cell and T-cell subsets (Table 1 and Suppl
160
Figures 1-5; reference values per age category are presented in Suppl Table 2).24, 33
AC C
EP
156
161
Hierarchical clustering of all 62 leukocyte populations resulted in 4 distinct clusters of
162
leukocyte dynamics (Figure 2A). Cell numbers of all populations in the first cluster were high
163
at birth, followed by a sharp decrease within the first 6 months of life, after which they
164
stabilized (Figure 2B). This cluster exclusively contained innate leukocyte subsets. All
165
defined innate leukocyte subsets clustered within cluster 1, except for the intermediate
166
monocytes.
10
ACCEPTED MANUSCRIPT The populations in cluster 2 strongly increased shortly after birth, peaked before 14
168
months of age, followed by a long-term gradual decrease. Cluster 2 included the 3 major
169
naive lymphocyte subsets (CD4+ Tnaive, CD8+ Tnaive and Bnaive), as well as 5 memory
170
populations (CD4+ central memory T cells (Tcm), CD8+ Tcm cells, early differentiated CD8+
171
CD45RO- effector memory T cells (CD8+ early TemRA), CD27-IgA+ memory B cells and
172
natural effector memory B cells). Also the total B-cell, CD8+ T-cells, CD4+ T-cells and total
173
T-cell subsets clustered in cluster 2, as well as the intermediate monocytes and the Vδ1+ T
174
cells.
M AN US C
RI PT
167
The populations in clusters 3 and 4 gradually increased in cell number and peaked either
176
at 14 months (cluster 3) or after ~4 years (cluster 4). Clusters 3 and 4 contained only memory
177
B-cell and T-cell subsets, and the TCRγδ+ T-cell subsets. Cluster 3 contained the total
178
memory lymphocyte populations Bmem, CD4+ Tmem and CD8+ Tmem. In particular, these
179
memory populations included the early, intermediate and late CD8+ TemRO subpopulations,
180
late CD8+ TemRA cells, early CD4+ TemRA cells, and the IgMonly, CD27+IgA+ and CD27-
181
IgG+ memory B-cell subsets. Cluster 4 contained total TCRγδ+ T cells and the Vδ2+ and the
182
Vγ9+ T-cell subsets, as well as the early, intermediate and late CD4+ TemRO subpopulations,
183
late CD4+ TemRA cells, intermediate CD8+ TemRA cells, and the CD27+IgG+ memory B
184
cells.
AC C
EP
TE
D
175
185
To test the effect of including populations with overlapping population definitions on
186
the clustering, an additional clustering was performed on a selection of 31 non-overlapping
187
populations (Table 1). The resulting clusters showed similar dynamics (Figure 2B and
188
Supplemental Figure 6B) with only 3 populations (non-classical monocytes, CD8+ late
189
TemRO and CD8+ late TemRA) being assigned to a different cluster, indicating that the
190
overlapping populations did not overtly skew our clustering approach. Thus, functionally
11
ACCEPTED MANUSCRIPT 191
related populations, as well as populations with phenotypic overlapping definitions displayed
192
similar dynamics in early childhood.
RI PT
193
Associations between nongenetic factors and leukocyte subset dynamics in the four
195
clusters
196
The memory B- and T-cell subset showed large inter-individual variation in dynamics of cell
197
numbers (Figure 1). To study the effects of external factors, we here analyzed 6 maternal life
198
style- and 3 maternal immune-related determinants, 6 birth characteristics, and 11
199
bacterial/viral exposure-related determinants (Table 2). Twelve of these 26 determinants
200
showed a significant association with the dynamics of one or more of the 62 leukocyte
201
populations after multivariable correction (Table 2). Bacterial/viral exposure-related
202
determinants were more frequently found to affect cell numbers in clusters 2, 3 and 4 than in
203
cluster 1 (Figure 3A-B).
TE
D
M AN US C
194
In cluster 1, a low maternal education level was associated with a reduction in the
205
patterns of eosinophils and classical monocytes (Figure 4). Female gender was associated
206
with a significant increase in the pattern of neutrophils, and the phenotypically overlapping
207
CD15+ granulocyte population. Having more than 1 sibling was associated with a reduction,
208
and CMV infection with an increase, in NK cells. Antibiotics usage in the first year of life
209
was associated with an increase in non-classical monocytes. Overall effect estimates and
210
associations of effects with specific age-intervals are presented in Suppl Table 3.
AC C
EP
204
211
In cluster 2, none of the determinants included in our study affected the patterns of more
212
than 25% of the populations (Figure 3C-D). However, female gender was associated with an
213
increase in the patterns of CD4+ naive and Tcm cells, and consequently in the phenotypically
214
overlapping CD4+TCRαβ+ and total CD4+ T cells. Whereas breastfeeding for more than 6
215
months was associated with a reduction in the pattern of total IgA+ and CD27-IgA+ memory
12
ACCEPTED MANUSCRIPT B cells, H. pylori carriership was associated with an increase in these populations.
217
Furthermore, H. pylori carriership was associated with an increase in total Igκ+ B cells and
218
TCRαβ+ T cells. CMV and EBV were both associated with an increase in the strongly related
219
total and TCRαβ+ CD8+ T cells, and CMV with an increase in Vδ1+ T cells. HSV-1
220
seropositivity was associated with a decrease in the large population of naive B cells, and
221
consequently with total B cells and Igκ+ and Igλ+ B-cell subsets.
M AN US C
RI PT
216
In cluster 3, CMV and EBV seropositivity were significantly associated with changes in
223
dynamics of >40% of subsets (Figure 3C; Suppl Table 3). Both viruses were associated with
224
an increase in CD8+ intermediate and late TemRO cells, and the phenotypically-related total
225
CD8+ TemRO and total CD8+ Tmem populations. In addition, CMV was associated with an
226
increase in CD8+TCRγδ+ T cells, and CD8+ late TemRA cells, and the phenotypically-related
227
total CD8+ TemRA population. EBV was associated with an increase in CD8+ early TemRO
228
cells, and a decrease in total Bmem, and the CD27- and CD27+ Bmem subsets. The effects on
229
CD8+ late TemRO and TemRA cells were still present in the analysis of 31 non-overlapping
230
population, even though these were now included in cluster 4 (Figure 3D).
EP
TE
D
222
In cluster 4, gender, premature rupture of membranes, CMV and EBV were significantly
232
associated with changes in dynamics of 40% of subsets (Figure 3C-D; Suppl Table 3).
233
Female gender was associated with an increase in the patterns of CD27+IgG+ memory B cells,
234
and CD4+ early TemRO cells, and with a reduction in total TCRγδ+ T cells and CD4+
235
intermediate and late TemRA populations. Premature rupture of membranes was associated
236
with an increase in CD4+ late TemRO, CD4+ intermediate TemRA and Vδ2+ and Vγ9+ T-cell
237
subsets. CMV and EBV seropositivity were both associated with an increase in total CD4+
238
TemRO and CD4+ late TemRO and TemRA cells, as well as CD8+ intermediate TemRA
239
cells. In addition, CMV was associated with increased CD4+ intermediate TemRO cells, and
240
EBV with an increase in CD4+ early TemRO cells.
AC C
231
13
ACCEPTED MANUSCRIPT Altogether, we identified multiple determinants that had a significant effect on the
242
dynamics of leukocyte subset numbers within one or more of the four distinct age-related
243
patterns. Still, considerable variation of effects could be observed within individual clusters.
RI PT
241
244
Effects of nongenetic factors on leukocyte dynamics within distinct leukocyte lineages
246
As clusters 2, 3 and 4 each contained various B-cell and T-cell populations, each with distinct
247
humoral or cellular functions, we next studied the effects of the 26 determinants on the
248
patterns of individual leukocyte lineages.
M AN US C
245
Within the B-cell lineage, low maternal educational level was associated with a reduction
250
in the patterns of total memory B cells (Figure 4). Female gender was associated with an
251
increase in the patterns of CD27+IgA+ and CD27+IgG+ memory B cells, though these effects
252
were not reflected in the total memory B-cell populations. Breastfeeding for more than 6
253
months was associated with a selective reduction in the pattern of CD27-IgA+ memory B
254
cells, antibiotics usage in the first year of life was associated with a selective increase in the
255
pattern of CD27-IgG+ memory B cells, and H. pylori carriership was associated with an
256
increase in the patterns of both CD27-IgA+ and CD27+IgA+ memory B cells. EBV infection
257
was associated with a reduction in the pattern of total memory B cells, though not with a
258
specific memory B-cell subset, HSV-1 infection was associated with a selective reduction in
259
the pattern of naive B cells.
AC C
EP
TE
D
249
260
Within the CD4+ and CD8+ T-cell lineage increased levels of serum anti-TPO was
261
associated with an increase in CD8+ Tcm cells. Female gender was associated with an
262
increase in the patterns of CD4+ naive, Tcm and early TemRO cells, although these effects
263
were not reflected in the total CD4+ memory T-cell populations. In contrast, female gender
264
was associated with a decrease in the patterns of CD4+ intermediate and late TemRA and
265
CD8+ late TemRA cells. Whereas premature rupture of membranes was associated with an
14
ACCEPTED MANUSCRIPT increase in CD4+ late TemRO and CD4+ intermediate TemRA cells, child birth via caesarian
267
section was associated with a reduction in the pattern of CD4+ early TemRO cells. Persistent
268
viral infection with CMV and EBV significantly affected CD4+ and CD8+ T-cell dynamics, in
269
contrast to breastfeeding, antibiotics usage, H. pylori, HSV-1 and VZV carriership. CMV
270
infection associated exclusively with an increase in the patterns of intermediate and late
271
differentiated CD4+ and CD8+ TemRO and TemRA cells. Associations with EBV were
272
slightly more variable, also including early differentiated CD4+ and CD8+ TemRO cells.
M AN US C
RI PT
266
Within the TCRγδ+ T-cell lineage, premature rupture of membranes was associated with
274
an increase in Vδ2+Vγ9+ T cells, CMV infection associated with an increase in the pattern of
275
Vδ1+ T cells.
AC C
EP
TE
D
273
15
ACCEPTED MANUSCRIPT DISCUSSION
277
We here modelled the kinetics of 62 leukocyte subsets, and identified distinct patterns of cell
278
numbers in the first 6 years of life for innate leukocytes, naive B and T cells, and memory B
279
and T cells. Unsupervised clustering revealed that leukocyte dynamics between birth and the
280
age of 6 years could be summarized into 4 major profiles, with either (1) early predominance
281
and fast decline to stable numbers, (2) gradual increased in first year followed by a gradual
282
decline, or a (3 and 4) slow increased in first 1-2 year followed by stabilization.
M AN US C
RI PT
276
Consistent with their early predominance, innate leukocyte kinetics was affected by
284
maternal education level, which might already influence the fetus prenatally. Low maternal
285
education level is strongly related to a less healthy life.34 These observations might suggests
286
that maternal life style is especially important for shaping of innate leukocyte populations,
287
although the exact mediator remains to be determined.
D
283
The non-classical and intermediate monocytes did not consistently cluster with the innate
289
leukocyte subsets due to a later peak in numbers or a slower decline. This altered kinetics
290
could be the result of these being derived from classical monocytes, and not directly from
291
precursors in bone marrow.35 Still, the various subsets of monocytes express distinct levels of
292
proteins involved in HLA-class II-dependent antigen presentation and CD40-CD40L co-
293
stimulation,35 and differ in parasite pattern recognition.35-38 Thus, our data support the need to
294
discriminate intermediate and non-classical monocytes from the dominant population of
295
classical monocytes in kinetics studies.
AC C
EP
TE
288
296
All naive B-cell and T-cell populations followed profile 2 with a gradual increase in the
297
first 14 months followed by a more gradual decline. After the first 1-2 years of life, B-
298
lymphocyte production in bone marrow decreases,39, 40 and the thymus starts to involute.41
299
These processes are likely causes of the decline we observed.
16
ACCEPTED MANUSCRIPT The memory B-cell populations showed different patterns. Natural effector and CD27-
301
IgA+ memory B-cell kinetics were similar to naive B cells. Both memory B-cell populations
302
are suggested to derive, at least in part, from germinal center-independent responses without
303
T-cell help in the splenic marginal zone and the intestinal lamina propria.33,
304
relatively fast rise in these cell numbers within the first 6 months of life contrasts the kinetics
305
of the other, T-cell dependent, memory B-cell subsets. IgMonly, CD27-IgG+ and CD27+IgA+
306
memory B cells followed profile 3, representing gradual memory formation. The CD27+IgG+
307
population followed profile 4, and developed slightly slower than the other memory B-cell
308
populations. These observations supports the concept of rapid generation of T-cell
309
independent memory B cells in the absence of the extensive proliferation and selection
310
processes of the germinal center,33 whereas CD27+IgG+ memory B cells might represent the
311
most mature memory B-cell population. In adults, this subset shows a higher level of affinity
312
maturation, more extensive replication history, and more frequent development via
313
consecutive class-switching, than CD27-IgG+ memory B cells.33, 44 It is perceivable that many
314
of these CD27+IgG+ cells are generated from CD27-IgG+ memory B cells in secondary
315
responses, which could explain the gradual decline in CD27-IgG+ B cell numbers after 14
316
months of age. Together, the different clustering of the memory B-cell populations thereby
317
follows their increasing functional maturity (Suppl Figure 5I).
RI PT
300
The
AC C
EP
TE
D
M AN US C
42, 43
318
Several external determinants significantly affected the B-cell kinetics. First, we
319
confirmed our previous observations regarding lower numbers of memory B cells in breastfed
320
children.20, 22 However, in the current analysis we could not reproduce the previously reported
321
association between breastfeeding duration and germinal center-dependent memory B-cell
322
numbers at 6 months of age.22 The difference is likely due to our current analysis in which
323
duration of breastfeeding was not included and the overall pattern between birth and 6 years
324
of age was studied.
17
ACCEPTED MANUSCRIPT IgA+ memory B cells were significantly increased in H. pylori positive children. The
326
colonization of the gastric mucus by H. pylori was found to correlate with an increase in total
327
blood B-cell counts,45 on top of the local expansion of antibody-secreting IgA+ cells,46 which
328
are important for the protection against H. pylori infection.47 Our observed expansion of
329
circulating IgA+ memory B cells included both the CD27+IgA+ and mucosa-derived CD27-
330
IgA+ subsets. This would suggest that the presence of H. pylori does not only result in a local
331
expansion of plasma cells, but also a systemic expansion of memory B cells in otherwise
332
asymptomatic carriers. It remains to be determined whether this expansion is beneficial for
333
the host as the bacterial protein CagA inhibits B-cell apoptosis and thereby increases the risk
334
for mucosa-associated B-cell malignancies.48, 49
M AN US C
RI PT
325
Memory T-cell subsets were found in clusters 2, 3 and 4. CD4+ and CD8+ Tcm (cluster
336
2) express lymph node homing markers and are the presumed precursors of effector memory
337
T cells.24, 50, 51 The early peak in Tcm numbers prior to those of effector memory T cells in
338
young children would fit with this function. Similarly, CD8+ early TemRA cells might be
339
precursors for further differentiated TemRA subsets. Vδ1+ T-cell numbers (cluster 2) peaked
340
prior to the Vδ2+ and Vγ9+ subsets (cluster 4), thereby confirming the previously observed
341
early shift from Vδ1+ to Vδ2+Vγ9+ predominance in children,52, 53
AC C
EP
TE
D
335
342
In line with previous observations in both children and elderly,22, 54-58 both CD4+ and
343
CD8+ memory T-cell dynamics were predominantly affected by CMV and/or EBV
344
seropositivity. This concerned mostly late effector memory T-cell numbers for CMV and
345
early effector memory T-cell subsets for EBV.24, 59 HSV-1 seropositivity did not affect CD4+
346
and CD8+ T-cell populations,22,
347
association has, to our knowledge, not been described before. EBV infection was associated
348
with a reduction in memory B cells, likely due to the selective EBV persistence in these
349
cells.60, 61 Our large-scale analysis allowed us to separate the herpesvirus-associated effects,
58
but was associated with a loss of naive B cells. This
18
ACCEPTED MANUSCRIPT 350
with CMV and EBV being associated with memory T-cell expansions, and HSV-1 and EBV,
351
with a decrease in naive or memory B-cell numbers, respectively. Gender had a widespread effect on 14 subsets within multiple leukocyte lineages.
353
Interestingly, girls showed a skewing of humoral and early differentiated CD4+ T-cell
354
responses over cellular cytotoxic responses, in contrast to boys. These effects might be
355
associated with differences in sex-hormone levels (testosterone, estradiol) that are already
356
detectable during early infancy, as well as with genetic differences between females and
357
males.62 These insight can especially be important for dissection of auto-immune diseases,
358
which are much more prevalent in females than in males, though predisposition for allergic
359
diseases seems to be opposite in infancy.62, 63
M AN US C
RI PT
352
360
Methodological considerations
362
The strength of this study is its prospective longitudinal population-based design with >1,000
363
children and the possibility to study 26 external determinants in 4 subgroups with
364
adjustments for major confounders. The linear mixed model approach enabled modeling of
365
cross-sectional data, with further improvement of the accuracy by incorporating additionally
366
available longitudinal follow-up data. Furthermore, we included only children with a two-
367
generation Dutch ancestry, which prevented interference of our analyses by strong ethnic and
368
cultural influences. However, extrapolation of our findings to different ethnic and cultural
369
populations might be limited, and would require additional analysis of ethnically-different
370
population cohorts.
AC C
EP
TE
D
361
371
The inclusion of 62 leukocyte populations allowed for the large-scale analysis of the
372
effect of external determinants on both total cell lineages as well as on small subsets defined
373
by extensive and detailed phenotypic definition. The overlap in some populations could have
374
skewed the hierarchical clustering. However, our selection of 31 phenotypically non-
19
ACCEPTED MANUSCRIPT overlapping populations resulted in clusters with similar patterns, indicating the robustness of
376
the 4 major patterns of leukocyte dynamics and the observed effects of external determinants.
377
Our study was primarily explorative with a focus on the identification of determinants
378
that affected leukocyte dynamics between birth and 6 years of age. Although we defined
379
whether determinants had a positive or negative association with leukocyte numbers, we were
380
unable to identify the exact nature of the effect, i.e. exactly when these effects presented and
381
whether these effects will be transient, persistent or potentially even increasing over time.
382
Consequently, more research into individual determinants will be needed to extend our
383
observations by specifying these effects.
M AN US C
RI PT
375
Finally, special consideration needs to be taken for serology of infectious agents that
385
were measured at the age of 6 years. We cannot determine the exact timing of primary
386
infections, and this can consequently be in the whole period preceding the age of 6 years.
387
Still, most H. pylori infections already occur in early childhood, and IgG seropositivity to
388
herpesviruses only appears several weeks to 3 months after infection. Thus, these
389
determinants were present already during or prior to the 6th year of life.
390
EP
TE
D
384
Conclusions
392
With our unbiased approach, we here, for the first time, classified leukocyte populations
393
according to their dynamics between birth and 6 years of age. Moreover, we identified
394
nongenetic factors that are associated with the dynamics of cell lineages or specific leukocyte
395
subsets and underlie, at least in part, the human immunological diversity. These newly
396
identified determinants can provide new targets for studies on the molecular processes that
397
regulate leukocyte development and immune responses, and that together underlie formation
398
of long-lasting immunity without inducing destructive, excessive or insufficient immune
399
responses.
AC C
391
20
ACCEPTED MANUSCRIPT ACKNOWLEDGEMENTS
401
This work was supported by an Erasmus MC Fellowship to M.C.v.Z, and was performed
402
within the framework of the Erasmus MC Postgraduate School Molecular Medicine. The
403
Generation R Study is conducted by the Erasmus MC, Erasmus University Rotterdam in
404
close collaboration with the School of Law and Faculty of Social Sciences of the Erasmus
405
University Rotterdam, the Municipal Health Service Rotterdam Metropolitan Area, the
406
Rotterdam Homecare Foundation, and the Stichting Trombosedienst & Artsenlaboratorium
407
Rijnmond. We gratefully acknowledge the contributions of children and parents, general
408
practitioners, hospitals, midwives and pharmacies in Rotterdam. We thank D. Zhao, K.A.M.
409
van Kester, M.A.W. Smits-te Nijenhuis, M.J Kolijn-Couwenberg and N.M.A. Nagtzaam for
410
technical support.
M AN US C
RI PT
400
D
411
CONFLICT OF INTEREST
413
All authors declare that no competing interests exist.
414
J.J.M.v.D., E.G.v.L., L.E.B-J, H.H., W.A.D., and M.C.v.Z. designed and supervised the flow
415
cytometry experiments; D.v.d.H., M.A.E.J. and H.B.-C. performed and analyzed most of the
416
experiments and contributed to data analyses; V.W.V.J. designed the Generation R Study;
417
H.A.M. designed the study, the data collection and the analyses of determinant data; K.N.
418
contributed to the statistical modeling; D.v.d.H., M.C.v.Z., and M.A.E.J. wrote the
419
manuscript; and all authors commented on the manuscript.
AC C
EP
TE
412
21
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
RI PT
REFERENCES 1. Comans-Bitter WM, de Groot R, van den Beemd R, Neijens HJ, Hop WC, Groeneveld K, et al. Immunophenotyping of blood lymphocytes in childhood. Reference values for lymphocyte subpopulations. J Pediatr 1997; 130:388-93. 2. van Gent R, van Tilburg CM, Nibbelke EE, Otto SA, Gaiser JF, Janssens-Korpela PL, et al. Refined characterization and reference values of the pediatric T- and B-cell compartments. Clin Immunol 2009; 133:95-107. 3. Tosato F, Bucciol G, Pantano G, Putti MC, Sanzari MC, Basso G, et al. Lymphocytes subsets reference values in childhood. Cytometry A 2015; 87:81-5. 4. Duchamp M, Sterlin D, Diabate A, Uring-Lambert B, Guerin-El Khourouj V, Le Mauff B, et al. B-cell subpopulations in children: National reference values. Immun Inflamm Dis 2014; 2:131-40. 5. Xanthou M. Leucocyte blood picture in healthy full-term and premature babies during neonatal period. Arch Dis Child 1970; 45:242-9. 6. Manroe BL, Weinberg AG, Rosenfeld CR, Browne R. The neonatal blood count in health and disease. I. Reference values for neutrophilic cells. J Pediatr 1979; 95:89-98. 7. Sagnia B, Ateba Ndongo F, Ndiang Moyo Tetang S, Ndongo Torimiro J, Cairo C, Domkam I, et al. Reference values of lymphocyte subsets in healthy, HIV-negative children in Cameroon. Clin Vaccine Immunol 2011; 18:790-5. 8. Huck K, Feyen O, Ghosh S, Beltz K, Bellert S, Niehues T. Memory B-cells in healthy and antibody-deficient children. Clin Immunol 2009; 131:50-9. 9. Morbach H, Eichhorn EM, Liese JG, Girschick HJ. Reference values for B cell subpopulations from infancy to adulthood. Clin Exp Immunol 2010; 162:271-9. 10. Piatosa B, Wolska-Kusnierz B, Pac M, Siewiera K, Galkowska E, Bernatowska E. B cell subsets in healthy children: reference values for evaluation of B cell maturation process in peripheral blood. Cytometry B Clin Cytom 2010; 78:372-81. 11. Driessen GJ, Dalm VA, van Hagen PM, Grashoff HA, Hartwig NG, van Rossum AM, et al. Common variable immunodeficiency and idiopathic primary hypogammaglobulinemia: two different conditions within the same disease spectrum. Haematologica 2013; 98:1617-23. 12. Brodin P, Jojic V, Gao T, Bhattacharya S, Angel CJ, Furman D, et al. Variation in the human immune system is largely driven by non-heritable influences. Cell 2015; 160:3747. 13. Orru V, Steri M, Sole G, Sidore C, Virdis F, Dei M, et al. Genetic variants regulating immune cell levels in health and disease. Cell 2013; 155:242-56. 14. Carr EJ, Dooley J, Garcia-Perez JE, Lagou V, Lee JC, Wouters C, et al. The cellular composition of the human immune system is shaped by age and cohabitation. Nat Immunol 2016; 17:461-8. 15. Pabst HF, Spady DW, Pilarski LM, Carson MM, Beeler JA, Krezolek MP. Differential modulation of the immune response by breast- or formula-feeding of infants. Acta Paediatr 1997; 86:1291-7. 16. Winkler B, Aulenbach J, Meyer T, Wiegering A, Eyrich M, Schlegel PG, et al. Formulafeeding is associated with shift towards Th1 cytokines. Eur J Nutr 2015; 54:129-38. 17. Hanson L, Silfverdal SA, Stromback L, Erling V, Zaman S, Olcen P, et al. The immunological role of breast feeding. Pediatr Allergy Immunol 2001; 12 Suppl 14:15-9. 18. Hanson LA, Korotkova M, Lundin S, Haversen L, Silfverdal SA, Mattsby-Baltzer I, et al. The transfer of immunity from mother to child. Ann N Y Acad Sci 2003; 987:199206.
22
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
RI PT
19. Duijts L, Ramadhani MK, Moll HA. Breastfeeding protects against infectious diseases during infancy in industrialized countries. A systematic review. Matern Child Nutr 2009; 5:199-210. 20. Hawkes JS, Neumann MA, Gibson RA. The effect of breast feeding on lymphocyte subpopulations in healthy term infants at 6 months of age. Pediatr Res 1999; 45:648-51. 21. M'Rabet L, Vos AP, Boehm G, Garssen J. Breast-feeding and its role in early development of the immune system in infants: consequences for health later in life. J Nutr 2008; 138:1782S-90S. 22. Jansen MA, van den Heuvel D, van Zelm MC, Jaddoe VW, Hofman A, de Jongste JC, et al. Decreased memory B cells and increased CD8 memory T cells in blood of breastfed children: the generation R study. PLoS One 2015; 10:e0126019. 23. Bekker V, Scherpbier H, Pajkrt D, Jurriaans S, Zaaijer H, Kuijpers TW. Persistent humoral immune defect in highly active antiretroviral therapy-treated children with HIV1 infection: loss of specific antibodies against attenuated vaccine strains and natural viral infection. Pediatrics 2006; 118:e315-22. 24. Appay V, van Lier RA, Sallusto F, Roederer M. Phenotype and function of human T lymphocyte subsets: consensus and issues. Cytometry A 2008; 73:975-83. 25. Jaddoe VW, van Duijn CM, Franco OH, van der Heijden AJ, van Iizendoorn MH, de Jongste JC, et al. The Generation R Study: design and cohort update 2012. Eur J Epidemiol 2012; 27:739-56. 26. Kruithof CJ, Kooijman MN, van Duijn CM, Franco OH, de Jongste JC, Klaver CC, et al. The Generation R Study: Biobank update 2015. Eur J Epidemiol 2014; 29:911-27. 27. Kalina T, Flores-Montero J, van der Velden VH, Martin-Ayuso M, Bottcher S, Ritgen M, et al. EuroFlow standardization of flow cytometer instrument settings and immunophenotyping protocols. Leukemia 2012; 26:1986-2010. 28. Korevaar TI, Medici M, de Rijke YB, Visser W, de Muinck Keizer-Schrama SM, Jaddoe VW, et al. Ethnic differences in maternal thyroid parameters during pregnancy: the Generation R study. J Clin Endocrinol Metab 2013; 98:3678-86. 29. Kiefte-de Jong JC, Jaddoe VW, Uitterlinden AG, Steegers EA, Willemsen SP, Hofman A, et al. Levels of antibodies against tissue transglutaminase during pregnancy are associated with reduced fetal weight and birth weight. Gastroenterology 2013; 144:72635 e2. 30. Jansen MA, Tromp, II, Kiefte-de Jong JC, Jaddoe VW, Hofman A, Escher JC, et al. Infant feeding and anti-tissue transglutaminase antibody concentrations in the Generation R Study. Am J Clin Nutr 2014; 100:1095-101. 31. den Hollander WJ, Holster IL, van Gilst B, van Vuuren AJ, Jaddoe VW, Hofman A, et al. Intergenerational reduction in Helicobacter pylori prevalence is similar between different ethnic groups living in a Western city. Gut 2014. 32. R Core Team; R Foundation for Statistical Computing V, Austria. R: A language and environment for statistical computing. . 2015. 33. Berkowska MA, Driessen GJ, Bikos V, Grosserichter-Wagener C, Stamatopoulos K, Cerutti A, et al. Human memory B cells originate from three distinct germinal centerdependent and -independent maturation pathways. Blood 2011; 118:2150-8. 34. Bouthoorn SH, Silva LM, Murray SE, Steegers EA, Jaddoe VW, Moll H, et al. Loweducated women have an increased risk of gestational diabetes mellitus: the Generation R Study. Acta Diabetol 2015; 52:445-52. 35. Wong KL, Tai JJ, Wong WC, Han H, Sem X, Yeap WH, et al. Gene expression profiling reveals the defining features of the classical, intermediate, and nonclassical human monocyte subsets. Blood 2011; 118:e16-31.
23
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
RI PT
36. Turner JD, Bourke CD, Meurs L, Mbow M, Dieye TN, Mboup S, et al. Circulating CD14brightCD16+ 'intermediate' monocytes exhibit enhanced parasite pattern recognition in human helminth infection. PLoS Negl Trop Dis 2014; 8:e2817. 37. Hijdra D, Vorselaars AD, Grutters JC, Claessen AM, Rijkers GT. Phenotypic characterization of human intermediate monocytes. Front Immunol 2013; 4:339. 38. Ziegler-Heitbrock L, Hofer TP. Toward a refined definition of monocyte subsets. Front Immunol 2013; 4:23. 39. Kogut I, Scholz JL, Cancro MP, Cambier JC. B cell maintenance and function in aging. Semin Immunol 2012; 24:342-9. 40. Jensen K, Schaffer L, Olstad OK, Bechensteen AG, Hellebostad M, Tjonnfjord GE, et al. Striking decrease in the total precursor B-cell compartment during early childhood as evidenced by flow cytometry and gene expression changes. Pediatr Hematol Oncol 2010; 27:31-45. 41. Flores KG, Li J, Sempowski GD, Haynes BF, Hale LP. Analysis of the human thymic perivascular space during aging. J Clin Invest 1999; 104:1031-9. 42. Weller S, Braun MC, Tan BK, Rosenwald A, Cordier C, Conley ME, et al. Human blood IgM "memory" B cells are circulating splenic marginal zone B cells harboring a prediversified immunoglobulin repertoire. Blood 2004; 104:3647-54. 43. Weller S, Faili A, Garcia C, Braun MC, Le Deist FF, de Saint Basile GG, et al. CD40CD40L independent Ig gene hypermutation suggests a second B cell diversification pathway in humans. Proc Natl Acad Sci U S A 2001; 98:1166-70. 44. Jackson KJ, Wang Y, Collins AM. Human immunoglobulin classes and subclasses show variability in VDJ gene mutation levels. Immunol Cell Biol 2014; 92:729-33. 45. Helmin-Basa A, Michalkiewicz J, Gackowska L, Kubiszewska I, Eljaszewicz A, Mierzwa G, et al. Pediatric Helicobacter pylori infection and circulating T-lymphocyte activation and differentiation. Helicobacter 2011; 16:27-35. 46. Mattsson A, Quiding-Jarbrink M, Lonroth H, Hamlet A, Ahlstedt I, Svennerholm A. Antibody-secreting cells in the stomachs of symptomatic and asymptomatic Helicobacter pylori-infected subjects. Infect Immun 1998; 66:2705-12. 47. Lee CK, Weltzin R, Thomas WD, Jr., Kleanthous H, Ermak TH, Soman G, et al. Oral immunization with recombinant Helicobacter pylori urease induces secretory IgA antibodies and protects mice from challenge with Helicobacter felis. J Infect Dis 1995; 172:161-72. 48. Umehara S, Higashi H, Ohnishi N, Asaka M, Hatakeyama M. Effects of Helicobacter pylori CagA protein on the growth and survival of B lymphocytes, the origin of MALT lymphoma. Oncogene 2003; 22:8337-42. 49. Lin WC, Tsai HF, Kuo SH, Wu MS, Lin CW, Hsu PI, et al. Translocation of Helicobacter pylori CagA into Human B lymphocytes, the origin of mucosa-associated lymphoid tissue lymphoma. Cancer Res 2010; 70:5740-8. 50. Sallusto F, Geginat J, Lanzavecchia A. Central memory and effector memory T cell subsets: function, generation, and maintenance. Annu Rev Immunol 2004; 22:745-63. 51. Sallusto F, Lenig D, Forster R, Lipp M, Lanzavecchia A. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 1999; 401:708-12. 52. Sandberg Y, Almeida J, Gonzalez M, Lima M, Barcena P, Szczepanski T, et al. TCRgammadelta+ large granular lymphocyte leukemias reflect the spectrum of normal antigen-selected TCRgammadelta+ T-cells. Leukemia 2006; 20:505-13. 53. Breit TM, Wolvers-Tettero IL, van Dongen JJ. Unique selection determinant in polyclonal V delta 2-J delta 1 junctional regions of human peripheral gamma delta T lymphocytes. J Immunol 1994; 152:2860-4. 24
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
RI PT
54. Bekker V, Bronke C, Scherpbier HJ, Weel JF, Jurriaans S, Wertheim-van Dillen PM, et al. Cytomegalovirus rather than HIV triggers the outgrowth of effector CD8+CD45RA+CD27- T cells in HIV-1-infected children. Aids 2005; 19:1025-34. 55. Weinberger B, Lazuardi L, Weiskirchner I, Keller M, Neuner C, Fischer KH, et al. Healthy aging and latent infection with CMV lead to distinct changes in CD8+ and CD4+ T-cell subsets in the elderly. Hum Immunol 2007; 68:86-90. 56. van Leeuwen EM, Remmerswaal EB, Vossen MT, Rowshani AT, Wertheim-van Dillen PM, van Lier RA, et al. Emergence of a CD4+CD28- granzyme B+, cytomegalovirusspecific T cell subset after recovery of primary cytomegalovirus infection. J Immunol 2004; 173:1834-41. 57. Chidrawar S, Khan N, Wei W, McLarnon A, Smith N, Nayak L, et al. Cytomegalovirusseropositivity has a profound influence on the magnitude of major lymphoid subsets within healthy individuals. Clin Exp Immunol 2009; 155:423-32. 58. Derhovanessian E, Maier AB, Hahnel K, Beck R, de Craen AJ, Slagboom EP, et al. Infection with cytomegalovirus but not herpes simplex virus induces the accumulation of late-differentiated CD4+ and CD8+ T-cells in humans. J Gen Virol 2011; 92:2746-56. 59. Appay V, Dunbar PR, Callan M, Klenerman P, Gillespie GM, Papagno L, et al. Memory CD8+ T cells vary in differentiation phenotype in different persistent virus infections. Nat Med 2002; 8:379-85. 60. Babcock GJ, Decker LL, Volk M, Thorley-Lawson DA. EBV persistence in memory B cells in vivo. Immunity 1998; 9:395-404. 61. Chaganti S, Heath EM, Bergler W, Kuo M, Buettner M, Niedobitek G, et al. EpsteinBarr virus colonization of tonsillar and peripheral blood B-cell subsets in primary infection and persistence. Blood 2009; 113:6372-81. 62. Ober C, Loisel DA, Gilad Y. Sex-specific genetic architecture of human disease. Nat Rev Genet 2008; 9:911-22. 63. Whitacre CC, Reingold SC, O'Looney PA. A gender gap in autoimmunity. Science 1999; 283:1277-8.
25
ACCEPTED MANUSCRIPT Table 1. Definition of leukocyte subsets
+ + + + + + + + + + + + + +
SSClowCD45+CD3+ SSClowCD3+TCRγδ-TCRαβ+ SSClowCD3+TCRγδ-TCRαβ+CD8-CD4+ SSClowCD3+TCRγδ-TCRαβ+CD8+CD4SSClowCD3+TCRγδ+TCRαβSSClowCD3+TCRγδ+TCRαβ-CD8-CD4+ SSClowCD3+TCRγδ+TCRαβ-CD8+CD4SSClowCD3+TCRαβ-Vδ1-Vδ2SSClowCD3+TCRαβ-Vδ1-Vδ2+ SSClowCD3+TCRαβ-Vδ1+Vδ2FSClowSSClowCD3+TCRαβ-Vγ9+ FSClowSSClowCD3+TCRαβ-Vδ2+Vγ9+ SSClowCD45+CD3+CD4+CD8FSClowSSClowCD3+CD8-CD197+CD45RO-CD27+CD28+ FSClowSSClowCD3+CD8- CD197- & CD197+CD45RO+ FSClowSSClowCD3+CD8-CD197+CD45RO+CD27+CD28+ FSClowSSClowCD3+CD8-CD197-CD45RO+ FSClowSSClowCD3+CD8-CD197-CD45RO+CD27+CD28+ FSClowSSClowCD3+CD8-CD197-CD45RO+CD27-CD28+ FSClowSSClowCD3+CD8-CD197-CD45RO+CD27-CD28FSClowSSClowCD3+CD8-CD197-CD45ROFSClowSSClowCD3+CD8-CD197-CD45RO-CD27+CD28+ FSClowSSClowCD3+CD8-CD197-CD45RO-CD27-CD28+ FSClowSSClowCD3+CD8-CD197-CD45RO-CD27-CD28SSClowCD45+CD3+CD4-CD8+ FSClowSSClowCD3+CD8+CD197+CD45RO-CD27+CD28+ FSClowSSClowCD3+CD8+ CD197- & CD197+CD45RO+ FSClowSSClowCD3+CD8+CD197+CD45RO+CD27+CD28+ FSClowSSClowCD3+CD8+CD197-CD45RO+ FSClowSSClowCD3+CD8+CD197-CD45RO+CD27+CD28+ FSClowSSClowCD3+CD8+CD197-CD45RO+CD27+CD28FSClowSSClowCD3+CD8+CD197-CD45RO+CD27-CD28FSClowSSClowCD3+CD8+CD197-CD45ROFSClowSSClowCD3+CD8+CD197-CD45RO-CD27+CD28+ FSClowSSClowCD3+CD8+CD197-CD45RO-CD27+CD28FSClowSSClowCD3+CD8+CD197-CD45RO-CD27-CD28SSClowCD45+CD19+ SSClowCD19+CD27-IgD+ SSClowCD19+ IgD- & CD27+IgD+ SSClowCD19+CD27-IgDSSClowCD19+CD27+IgDSSClowCD19+CD27+IgM+ SSClowCD19+CD27+IgM+IgD+ SSClowCD19+CD27+IgM+IgDSSClowCD19+IgA+ SSClowCD19+CD27-IgA+ SSClowCD19+CD27+IgA+ SSClowCD19+IgG+ SSClowCD19+CD27-IgG+ SSClowCD19+CD27+IgG+ SSClowCD19+CD38lowCD21low SSClowCD19+Igκ-Igλ+ SSClowCD19+Igκ+Igλ-
+ + + +
+ + + + + +
RI PT
Lymphocytes T cells TCRαβ+ T cells ∟CD4+ TCRαβ+ T cells ∟CD8+ TCRαβ+ T cells TCRγδ+ T cells ∟CD4+ TCRγδ+ T cells ∟CD8+ TCRγδ+ T cells ∟Vδ1-Vδ2- TCRαβ- T cells ∟Vδ2+ T cells ∟Vδ1+ T cells ∟Vγ9+ T cells ∟Vδ2+Vγ9+ T cells CD4+ T cells ∟CD4+ Tnaive ∟CD4+ Tmem ∟CD4+ Tcm ∟CD4+ TemRO ∟CD4+ early TemRO ∟ CD4+ interm TemRO ∟CD4+ late TemRO ∟CD4+ TemRA ∟CD4+ early TemRA ∟ CD4+ interm TemRA ∟CD4+ late TemRA CD8+ T cells ∟CD8+ Tnaive ∟CD8+ Tmem ∟CD8+ Tcm ∟CD8+ TemRO ∟CD8+ early TemRO ∟CD8+ interm TemRO ∟CD8+ late TemRO ∟CD8+ TemRA ∟CD8+ early TemRA ∟CD8+ interm TemRA ∟CD8+ late TemRA B cells Bnaive Bmem ∟CD27- Bmem ∟CD27+ Bmem ∟IgM+ Bmem ∟Natural effector ∟IgMonly ∟IgA+ Bmem ∟CD27-IgA+ ∟CD27+IgA+ ∟IgG+ Bmem ∟CD27-IgG+ ∟CD27+IgG+ CD21low B cells Igλ+ B cells Igκ+ B cells
M AN US C
+
SSChighCD45+ SSChighCD45+CD15+ SSChighCD45+CD16+ SSChighCD45+CD16FSCinterSSCinterCD45+ FSCinterSSCinterCD45+CD14+CD16FSCinterSSCinterCD45+CD14+CD16+ FSCinterSSCinterCD45+CD14-CD16+ SSClowCD45+CD3-CD16+or CD56+
D
+ + + +
Phenotype definition
TE
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62
+ +
Name population Innate leukocytes Granulocytes CD15+ granulocytes Neutrophils Eosinophils Monocytes Classical monocytes Intermediate monocytes Non-classical monocytes NK cells
EP
1 2 3 4 5 6 7 8 9
n=31
AC C
n=62
26
Table 2. Characteristics of nongenetic factors negative individuals (%)
IP T
ACCEPTED MANUSCRIPT
positive individuals (%)
missing (%)
significant effect on cluster a
AC C
EP
TE
D
M AN US
CR
Prenatal maternal life style Maternal age (years) (continuous determinant) 0 (0) Low maternal educational level 400 (33.8) 760 (64.3) 22 (1.9) 1,2,3,4 Net household income per month >€ 2,200 242 (20.5) 847 (71.7) 93 (7.9) Smoking during pregnancy 815 (69) 256 (21.7) 111 (9.4) Alcohol use continued during pregnancy 299 (25.3) 750 (63.5) 133 (11.3) Body Mass Index before pregnancy (kg/m2) (continuous determinant) 181 (15.3) Prenatal maternal immune-mediated diseases Maternal atopy (eczema, allergy HDM, hay-fever 683 (57.8) 383 (32.4) 116 (9.8) Anti-TPO (before 18 weeks of pregnancy) (mU/ml) (continuous determinant) 278 (23.5) 2 Maternal TG2A during pregnancy U/ml (continuous determinant) 208 (17.6) Birth characteristics Gender (girl yes/no) 600 (50.8) 582 (49.2) 0 (0) 1,2,3,4 Preterm birth <37 weeks 76 (6.4) 1,106 (93.6) 0 (0) Low birth weight <2,500g 1,119 (94.7) 63 (5.3) 0 (0) Premature rupture of membranes 1,089 (92.1) 53 (4.5) 40 (3.4) 4 Caesarian section versus vaginal/ forceps/vacuum assisted 903 (76.4) 154 (13) 125 (10.6) 3,4 Birth season (born in Fall/Winter) 682 (57.7) 500 (42.3) 0 (0) Bacterial/viral exposure-related characteristics Breastfeeding ever 109 (9.2) 969 (82) 104 (8.8) Breastfeeding at 6 months of age 754 (63.8) 299 (25.3) 129 (10.9) 2,4 Siblings >1 687 (58.1) 480 (40.6) 15 (1.3) 1 Antibiotics/Penicillin use in 1st y 636 (53.8) 342 (28.9) 204 (17.3) 1,3 Upper respiratory tract infections in 1st y 547 (46.3) 475 (40.2) 160 (13.5) Lower respiratory tract infections in 1st y (doctor attended) 819 (69.3) 131 (11.1) 232 (19.6) Carrier of Helicobacter pylori within 6yrs 887 (75) 49 (4.1) 246 (20.8) 2,3 Seropositivity for Cytomegalovirus (CMV) at 6y 665 (56.3) 269 (22.8) 248 (21) 1,2,3,4 Seropositivity for Epstein Barr virus (EBV) at 6y 534 (45.2) 400 (33.8) 248 (21) 2,3,4 Seropositivity for Herpes simplex virus-1 (HSV-1) at 6y 810 (68.5) 124 (10.5) 248 (21) 2 Seropositivity for Varicella zoster virus (VZV) at 6y 73 (6.2) 861 (72.8) 248 (21) a First, each determinant was added to the linear mixed effect model univariably to test for a significant effect (Significance was defined as p<0.0125); Subsequently, per leukocyte population, all determinants with a significant individual effect were added in a multivariable model to correct for confounding effects. Up to 3 determinants were added per multivariable model. Significance was defined as p<0.05. IU, international units. No confounding effect was observed
27
ACCEPTED MANUSCRIPT Figure Legends (n=4)
in children between birth and 6 years of age.
RI PT
Figure 1. Dynamics of innate leukocyte and naive and memory lymphocyte populations
Dynamics of A) monocytes, neutrophils, eosinophils and NK cells, B) naive and memory B-
M AN US C
cell and T-cell subsets, and C) TCRγδ+ T cells. Flow cytometry plots depict population definitions in one representative 6-year-old individual. Graphs depict absolute numbers of cells in blood of 1,182 children with in total 2,010 measurements (gray dots). Linear mixed effect models are depicted by a solid black line for each population and the 90% confidence interval (CI) of the model with dashed black lines. For clarity, only direct follow-up time
D
points of an individual were connected with gray lines; i.e. 0-6m, 6-14m, 14-25m or 25-76m.
TE
Figure 2. Hierarchical clustering of leukocyte subset dynamics in early childhood. A) Ward’s hierarchical clustering of the cellularity of the 62 populations derived from the
EP
linear mixed models. Cellularity between 0-76 months of age was converted to zero mean
AC C
and unit standard deviation (z score). B) Average patterns +/- 1 standard deviation of the subsets in each of the 4 major clusters. Populations are numbered according to their order in Table 1.
Figure 3. Nongenetic factors associated with leukocyte dynamics in the four clusters. Significant associations of the 26 determinants with leukocyte subsets in clusters 1-4 based on the clustering of 62 leukocyte populations as in Figure 2 (A and C) or of the 31 nonoverlapping subsets as in Suppl Figure 6 (B and D) corrected for confounding effects. A and B) Relative distribution of types of determinants that significantly affect leukocyte kinetics. C and D) Frequencies of populations within a cluster that were affected by individual 28
ACCEPTED MANUSCRIPT determinants. Abbreviations: PROMM, premature rupture of membranes; RTI, respiratory
RI PT
tract infection. Numbers in each bar refer to the populations as defined in Table 1.
Figure 4. Effects of nongenetic factors on the dynamics of innate leukocytes, and B-cell and T-cell subsets.
M AN US C
Significant effects are shown derived from multivariable linear mixed effect analysis of all 26 determinants on the 31 non-overlapping populations, and the total memory subsets (“mem”) within the B-cell, CD4+ T-cell and CD8+ T-cell lineages. Blue boxes represent a reduction, red boxes an increase in the pattern of the indicated population. Effect sizes are presented in Suppl Table 3, and details of each determinant in Table 2. Additional abbreviations: PROMM, premature rupture of membranes; RTI, respiratory tract infection; neu, neutrophils;
D
eos, eosinophils; class, classical monocytes; inter, intermediate monocytes; n-class, non-
AC C
EP
TE
classical monocytes.
29
B lympho
monocytes
3000 2000
Tnaive
2000 1000 500 0
0
D
age (months)
Tnaive
cells/µl
CD45RO
TE
12 24 36 48 60 72 84 96
AC C
CD3
T cells
TCRγδ+ T cells
1500
1000 500
12 24 36 48 60 72 84 96
CD4+ Tnaive
0
2500
12 24 36 48 60 72 84 96
CD4+ Tmem
2000 1500 1000 500 0
0
12 24 36 48 60 72 84 96
age (months)
1000 800 600 400 200
0
CD8+ Tmem
2000
0
0
9000 8000 7000 6000 5000 4000 3000 2000 1000 0
0
TCRαβ
12 24 36 48 60 72 84 96
1500
1000
TCRγδ+ T cells
1200
0
1500
1400
cells/μl
2500
C
EP
NK cells
3000
Tmem
CCR7
12 24 36 48 60 72 84 96
3500
cells/µl
NK cells
CD4+
CD3
0
lymphocytes
CCR7 CD4+ T cells
CD8+
TCRγδ
CD45
eosinophils
CD8
eosino
9000 8000 7000 6000 5000 4000 3000 2000 1000 0
CD8+ Tnaive
2000
0
12 24 36 48 60 72 84 96
Bmem
2500
500
5000
cells/µl
CD16
T-cells
10000
Tmem
cells/µl
cells/µl
15000
0
12 24 36 48 60 72 84 96
3000 2500
neutrophils
20000
0
2000
0
CD27 CD8+ T cells
CD19
12 24 36 48 60 72 84 96
30000
neutro
3000
0 0
0 25000
granulocytes
800 700 600 500 400 300 200 100 0
Bnaive
1000
1000 0
CD45
CD16/CD56
B cells
1500 500
lympho
5000 4000
CD45RO
mono
2500
CD3
SSC
cells/µl
granulo
B cells Bnaive Bmem
T cells
cells/µl
3500
IgD
all events
M AN U
A
SC
RI PT
ACCEPTED MANUSCRIPT
12 24 36 48 60 72 84 96
age (months)
0
12 24 36 48 60 72 84 96
age (months)
Data of individual children model 90% CI
3
4
0
M AN U
2
SC
4 1 2 3 9 6 5 8 25 52 51 55 62 46 34 47 61 60 13 43 19 54 7 37 17 15 23 10 11 35 22 12 40 38 39 36 41 50 45 42 16 53 24 56 48 58 49 57 31 30 18 28 44 14 32 21 33 20 27 26 29 59
1
10
20
30
D
A
RI PT
ACCEPTED MANUSCRIPT
40
50
Age (months)
TE -2
Z- score
1
0
3.00 2.50 2.00 1.50 1.00 0.50 0.00 -0.50 -1.00 -1.50 -2.00 3.00 2.00 1.00 0.00 -1.00 -2.00 -3.00 -4.00 -5.00
6
2
0
0
1
2
Eosinophils Granulocytes CD15+ granuloytes Neutrophils NK cells Classical monocytes Monocytes Non-classical monocytes
6
12 18 24 30 36 42 48 54 60 66 72
12 18 24 30 36 42 48 54 60 66 72
3
0
2.00
6
4
0.00 -1.00 -2.00 -3.00 -4.00 -5.00
0
6
25 52 51 55 62 46 34 47 61 60 13 43
CD4+ Tcm Natural effector B cells IgM+ Bmem CD27-IgA+ Igκ+ B cells B cells CD8+ T cells Bnaive Igλ+ B cells CD21low B cells CD8+ TCRαβ+ T cells CD8+ TemRA early
19 54 7 37 17 15 23 10 11 35 22 12
Vδ1+ T cells IgA+ Bmem Intermediate monocytes CD8+ Tcm Vδ1-Vδ2- TCRαβ- T cells CD4+ TCRγδ+ T cells CD4+ Tnaive T cells TCRαβ+ T cells CD8+ Tnaive CD4+ T cells CD4+ TCRαβ+ T cells
40 38 39 36 41 50 45 42 16
CD8+ TemRO interm CD8+ TemRO CD8+ TemRO early CD8+ Tmem CD8+ TemRO late CD27+ Bmem CD8+ TemRA late CD8+ TemRA CD8+ TCRγδ+ T cells
53 24 56 48 58 49 57 31 30
IgMonly CD4+ Tmem CD27+IgA+ Bmem CD27-IgG+ CD27- Bmem IgG+ Bmem CD4+ TemRA early CD4+ TemRA
18 28 44 14 32 21 33 20 27
Vδ2+ T cells CD4+ TemRO interm CD8+ TemRA interm TCRγδ+ T cells CD4+ TemRA interm Vδ2+Vγ9+ T cells CD4+ TemRA late Vγ9+ T cells CD4+ TemRO early
26 CD4+ TemRO 29 CD4+ TemRO late 59 CD27+IgG+
12 18 24 30 36 42 48 54 60 66 72
1.00
Z- score
70
EP
4 1 2 3 9 6 5 8
-1
AC C
7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00 -1.00 -2.00
Z- score
Z- score
B
60
z-score
12 18 24 30 36 42 48 54 60 66 72 Age (months)
A
B
Total 62 populations
31 non-overlapping populations cluster 1
(n=8 effects)
(n=5 effects)
TE
cluster 1 cluster 2
cluster 2
(n= 21 effects)
(n=8 effects)
cluster 3
cluster 3
(n=6 effects)
(n= 24 effects)
EP
cluster 4
cluster 4
(n=20 effects)
10
20
30
40
50
60
70
80
90 100
0
distribution of determinants with a significant effect (%)
Total 62 populations Cluster 1 8 leukocyte pop. 4 5 6
2 3
9
Cluster 2 24 leukocyte pop.
60
37
12 22 23 25
8
9
11 54 55 62
54 55
13 19 34
13 34
16 36 38 40 41 42 56 45
36 38 39 40 41 48 49 50
Cluster 3 18 leukocyte pop.
16 48 49
42 45 56
58
16
46 47 61 62
Cluster 4 12 leukocyte pop. 14 27 32 33 59
18
18 20 21 29 32
26 28 29 33 44 27
26 27 29 33 44
18
age education household income smoking alcohol BMI atopy TPO tTG gender preterm birth low birth weight PROMM sectio birth season breastfeeding ever breastfeeding >6m siblings antibiotics 1y upper RTI 1y lower RTI 1y H.pylori CMV EBV HSV-1 VZV
frequency of populations affected within each cluster (%)
50 40 30 20 10 0 50 40 30 20 10 0 50 40 30 20 10 0 50 40 30 20 10 0
prenatal maternal life style
immunemediated diseases
birth charact.
20
birth characteristics
AC C
maternal immune-mediated
life style
10
30
40
50
60
70
80
90 100
distribution of determinants with a significant effect (%)
bacterial/viral exposure-related characteristics
bacterial/viral exposure
D
31 non-overlapping populations 60 50 40 30 4 20 6 10 0 60 50 40 30 20 10 0 60 50 40 30 20 10 0 60 50 40 30 20 10 0
Cluster 1 4 leukocyte pop.
3
9
9
Cluster 2 11 leukocyte pop.
23 37 25
55
8 55 19
47
Cluster 3 6 leukocyte pop.
56
58 56 40
39 40
Cluster 4 10 leukocyte pop. 27 32 33 21 45 29 59 32 27
28 29 33 41 44 45
27 29 33 41 44.
education TPO gender PROMM sectio breastfeeding >6m siblings antibiotics 1y H.pylori CMV EBV HSV-1
0
frequency of populations affected within each cluster (%)
(n= 23 effects)
C
D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
granulo neu
SC M AN U D
bacterial/viral exposure
TE
birth charact.
age education household income smoking alcohol BMI atopy TPO tTG gender preterm birth low birth weight PROMM sectio birth season breastfeeding ever breastfeeding >6m siblings antibiotics 1y upper RTI 1y lower RTI 1y H. pylori CMV EBV HSV-1 VZV
EP
immunemediated diseases
CD8+ T cells TCRγδ B cells CD4+ T cells IgA+ IgG+ TemRA TemRO TemRO TemRA Nat. IgM Vδ2 NK naive mem eff. only CD27-CD27+ CD27-CD27+ naive mem Tcm early interm late early interm late naive mem Tcm early interm late early interm late Vδ1 Vγ9
AC C
life style
Innate monocytes
eos class inter n-class
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
IP T
Supplemental Table 1. Antibody details
Fluorochrome FITC PE PerCP-Cy5.5 PE-Cy7 APC APC-Cy7 Antibody CD3 CD16+CD56 CD45 CD4 CD19 CD8 Clone SK7 B73.1 + C5.9 2D1 SK3 SJ25C1 SK1 1 Manufacturer BD BD+Dako BD BD BD BD Antibody CD15 CD45 CD16 CD14 Clone MMA 2D1 3G8 MO-P9 2 Manufacturer BD BD BD BD Antibody Igκ Igλ CD19 CD21 Clone polyclonal polyclonal SJ25C1 B-ly-4 3 Manufacturer Dako Southern Biotech BD BD Antibody CD38 CD19 CD21 Clone HB7 SJ25C1 B-ly-4 4 Manufacturer BD BD BD Antibody IgD CD19 IgM CD27 Clone polyclonal SJ25C1 polyclonal L128 5 Manufacturer Southern Biotech BD BD BD Antibody IgA IgG CD19 IgM CD27 Clone polyclonal polyclonal SJ25C1 polyclonal L128 6 Manufacturer Kallestad Southern Biotech BD BD BD Antibody TCRαβ TCRγδ CD3 CD4 CD8 Clone WT31 11F2 SK7 SK3 SK1 7 Manufacturer BD BD BD BD BD Antibody Vδ2 Vδ1 * CD3 CD4 TCRαβ CD8 Clone B6.1 R9.12 SK7 SK3 IP26 SK1 8 Manufacturer BD Beckman Coulter BD BD eBiosciences BD Antibody CD28 CD197 CD3 CD8 CD45RO CD27 Clone CD28.2 3D13 SK7 SK1 UCHL-1 L128 9 Manufacturer BD eBiosciences BD BD BD BD Antibody Vδ2 Vγ9 CD3 CD4 TCRαβ CD8 Clone B6.1 B3.1 SK7 SK3 IP26 SK1 10 Manufacturer BD BD BD BD eBiosciences BD *, unconjugated antibody, detected with Goat anti-Mouse IgG PE (polyclonal; Invitrogen, Waltham, MA) BD, BD Biosciences, San Jose, CA; Dako, DakoCytomation, Glostrup, Denmark; Southern Biotech, Southern Biotech, Birmingham, AL; Kallestad, Kallestad Diagnostics, Chaska, MN; Beckman Coulter, Beckman Coulter, Indianapolis, IN; eBiosciences, eBiosciences, San Diego, CA.
AC C
EP
TE
D
M AN US
CR
Tube
ACCEPTED MANUSCRIPT Supplemental Table 2. Age-associated reference values of leukocyte populations
Age (months)
0 220 0.0 0.0 0.0 0.0 0.0
5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th
4884.2 8263.0 10572.7 13727.9 18784.8 5161.2 8010.4 10397.8 13564.7 17933.3 4433.3 7064.5 9091.8 11634.0 16483.1 489.5 891.3 1340.0 1999.2 3721.9 496.1 1002.2 1346.5 1749.6 2429.0 378.1 829.5 1114.3 1425.4 2125.5 20.6 46.6 72.3 112.1 170.0 24.0 50.6 77.3 116.0 207.5 229.5 460.0 840.0 1350.0 2100.5
Age category (months) 6 14 25 376 241 257 5.5 13.6 23.6 6.0 14.1 24.4 6.2 14.4 25.1 6.7 14.8 25.8 7.7 16.1 27.3
Innate Leukocytes
Neutrophils
EP
AC C
Monocytes
TE
Eosinophils
Classical monocytes
Intermediate monocytes
Non-classical monocytes
NK cells
1196.3 2250.4 3074.1 4143.0 6056.8 1138.2 2072.7 2973.2 4066.6 5633.9 1026.1 1926.4 2750.9 3795.5 5506.1 84.5 178.1 248.0 373.2 763.6 320.6 504.0 691.6 962.7 1510.6 192.7 301.7 425.3 572.6 849.1 26.1 42.4 70.3 116.7 243.3 13.4 24.3 41.3 66.1 139.2 130.0 210.0 280.0 400.0 690.0
1352.5 1990.9 2822.9 3832.6 6467.5 1310.8 1931.9 2690.5 3746.3 6194.1 1116.3 1840.5 2533.5 3665.9 5983.6 83.9 142.2 203.8 315.6 617.5 272.8 433.1 577.4 729.6 1066.6 190.0 304.0 402.6 522.9 714.1 11.6 27.9 43.7 72.0 124.3 4.2 12.6 26.8 48.9 93.4 120.0 180.0 240.0 320.0 536.0
M AN US C
CD15+ granulocytes
D
Granulocytes
1026.1 2055.0 2850.0 4167.8 6519.8 871.6 1735.8 2617.3 3789.1 5913.7 824.4 1668.6 2497.6 3671.1 5653.1 155.2 253.1 351.6 486.7 856.0 360.1 582.8 789.6 1065.8 1755.0 220.4 338.6 461.1 623.9 992.3 19.4 36.0 55.1 87.0 202.7 16.2 39.3 60.8 92.6 162.2 160.0 250.0 330.0 450.0 752.5
72 916 69.3 70.6 72.4 75.0 79.0
RI PT
Number of children
percentile 5th 25th 50th 75th 95th
1562.4 2840.2 3715.0 5041.0 7597.8 1521.0 2736.1 3608.8 4896.7 7527.4 1321.7 2414.2 3274.0 4451.2 6810.9 109.4 199.1 356.4 598.1 1261.5 233.9 370.4 490.9 623.7 924.1 168.8 267.4 349.6 455.9 676.4 10.1 21.0 31.0 44.8 89.8 4.8 11.7 22.0 41.2 80.3 90.0 140.0 190.0 250.3 420.0
ACCEPTED MANUSCRIPT
CD4+TCRαβ+ T cells
CD8+TCRαβ+ T cells
TCRγδ+ T cells
AC C
CD8+ TCRγδ+ T cells
Vδ1-Vδ2- TCRαβ- T cells
Vδ2+ T cells
Vδ1+ T cells
Vγ9+ T cells
2587.5 3740.0 4620.0 5690.0 7722.5 2509.1 3680.3 4477.9 5426.5 7552.0 1788.3 2679.3 3295.1 3994.7 5597.1 575.1 856.5 1101.7 1427.4 2112.0 69.1 123.3 164.0 226.5 333.2 6.6 13.7 20.4 31.6 50.8 8.5 18.2 27.7 44.4 81.0 69.1 113.5 173.8 274.8 448.2 13.1 30.4 43.5 65.5 124.4 29.0 48.2 65.1 96.4 163.8 25.0 39.9 56.7 80.8 119.0
2180.0 3080.0 3700.0 4610.0 6010.0 2063.6 2934.0 3534.7 4401.4 5697.5 1388.1 1926.2 2401.3 3023.6 4019.3 503.3 796.1 1013.3 1291.6 1894.5 68.8 106.5 162.8 213.2 332.6 4.8 9.1 13.3 19.3 36.0 7.2 15.7 28.3 42.6 92.5 28.0 53.6 89.9 130.6 288.3 11.4 28.8 47.5 74.1 127.8 23.4 42.2 62.8 87.4 135.5 28.0 43.5 62.8 100.3 181.9
1586.0 2220.0 2830.0 3650.0 4956.0 1461.9 2064.0 2691.3 3345.1 4565.6 896.6 1338.0 1739.1 2233.5 3167.0 411.9 630.5 813.2 1053.7 1582.7 63.7 109.3 163.1 219.8 352.8 3.4 6.6 10.4 14.7 25.2 6.7 14.3 25.7 41.2 75.4 30.6 49.0 70.7 95.9 151.5 12.9 30.8 53.5 90.9 177.5 18.5 33.2 49.7 67.8 134.4 24.7 45.1 67.1 99.2 174.8
72 1329.0 1770.0 2102.0 2571.8 3560.0 1170.2 1593.1 1916.9 2328.9 3295.1 685.2 949.5 1177.0 1450.3 2067.5 364.1 537.4 669.9 857.0 1227.1 82.3 132.2 178.2 248.2 377.7 2.1 4.4 6.9 10.4 18.5 6.9 14.3 22.3 33.7 62.2 23.6 43.9 65.9 93.6 164.7 22.8 50.9 77.1 118.4 212.9 13.6 25.4 37.7 54.3 94.4 25.5 46.8 69.5 104.0 187.0
RI PT
1478.5 2187.5 2695.0 3457.5 4745.5 1403.3 2130.3 2607.5 3318.6 4598.5 957.5 1446.6 1848.8 2338.7 3085.2 351.4 542.7 736.4 957.6 1454.5 31.5 63.1 87.7 126.0 225.1 5.4 11.5 17.6 26.0 43.5 3.2 7.6 12.6 19.4 39.8 22.9 43.1 62.5 89.7 164.1 4.0 8.1 12.2 19.7 35.0 7.6 18.6 30.7 43.8 69.2 15.1 20.1 29.7 38.7 53.6
EP
CD4+ TCRγδ+ T cells
5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th
M AN US C
TCRαβ+ T cells
0
D
T cells
Age category (months) 6 14 25
percentile
TE
Supplemental Table 2 (continued) Lymphocytes
ACCEPTED MANUSCRIPT
CD4+ Tnaive
CD4+ Tmem
CD4+ Tcm
EP
TE
CD4+ TemRO
AC C
CD4+ early TemRO
CD4+ intermediate TemRO
CD4+ late TemRO
CD4+ TemRA
CD4+ early TemRA
Age category (months) 6 14 25 5.2 9.0 8.7 11.8 18.3 20.9 22.9 31.7 39.2 38.4 58.0 66.0 71.0 116.5 135.7 1820.5 1430.2 922.3 2706.7 1991.4 1362.5 3366.2 2457.8 1782.7 4101.9 3062.0 2288.7 5553.1 4123.0 3217.7 1519.9 941.0 575.3 2233.6 1469.5 916.6 2849.1 1972.0 1256.0 3460.9 2461.5 1662.9 4841.4 3373.4 2336.6 354.1 331.6 325.1 470.8 526.9 504.6 580.2 669.0 651.4 713.2 834.9 850.9 953.0 1227.9 1274.9 134.9 111.0 86.7 202.7 176.0 129.5 261.9 242.8 173.1 322.1 313.2 223.6 449.3 427.4 349.3 61.2 31.8 76.9 93.0 123.7 152.4 138.3 184.3 205.4 177.5 254.7 260.3 260.8 387.9 387.4 50.8 20.6 54.6 79.0 95.6 116.7 110.3 148.8 154.8 144.8 199.8 197.5 211.9 301.6 276.4 2.7 2.6 9.7 10.2 14.0 17.7 17.3 23.0 26.2 24.7 33.2 36.4 38.5 53.4 64.0 0.0 0.0 1.4 0.0 2.5 3.8 0.0 5.1 7.7 3.8 11.2 17.6 14.3 29.0 43.4 79.9 50.3 58.1 122.0 115.1 144.0 178.0 221.2 245.8 241.2 343.4 415.3 376.8 570.5 812.5 66.7 33.1 40.3 104.3 91.1 114.3 149.9 189.8 213.6 204.6 304.4 356.7 357.8 542.8 782.6
72 13.3 31.0 49.9 83.9 163.1 737.1 988.2 1240.0 1525.3 2134.0 392.9 621.4 821.9 1069.7 1538.0 344.5 461.7 559.2 679.5 878.5 74.2 112.5 149.0 198.9 285.0 118.3 181.5 235.2 295.5 408.5 82.5 127.1 168.0 210.1 292.4 18.2 29.7 43.9 61.5 105.2 1.1 3.0 6.3 12.9 30.1 55.8 102.4 153.6 227.8 371.7 36.3 67.6 110.7 168.4 317.4
RI PT
CD4+ T cells
0 1.5 3.6 5.7 12.5 24.3 1060.7 1552.3 1972.5 2498.9 3266.2 848.7 1321.1 1695.9 2159.7 2898.2 161.0 259.6 355.9 465.1 754.6 39.5 81.4 121.3 198.3 395.6 23.4 51.5 73.8 107.0 172.6 21.0 47.6 65.0 98.9 155.8 0.0 1.8 3.8 7.0 19.6 0.0 0.0 0.0 0.0 3.3 35.9 89.2 126.8 181.1 312.1 27.9 74.2 113.7 162.7 287.3
M AN US C
Vδ2+Vγ9+ T cells
percentile 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th
D
Supplemental Table 2 (continued)
ACCEPTED MANUSCRIPT
CD8+ T cells
CD8+ Tnaive
CD8+ Tmem
EP
AC C
CD8+ TemRO
TE
CD8+ Tcm
CD8+ early TemRO
CD8+ intermediate TemRO
CD8+ late TemRO
CD8+ TemRA
Age category (months) 6 14 25 0.0 0.0 0.0 0.0 0.0 1.4 3.6 2.7 2.7 5.7 4.6 4.7 11.0 9.0 7.8 0.0 0.0 1.7 0.0 3.3 4.4 3.6 6.9 8.8 6.6 15.2 16.0 17.8 44.5 42.0 527.3 447.2 389.3 836.6 776.2 619.4 1093.1 998.8 805.1 1389.5 1277.1 1038.8 2060.9 1831.8 1571.6 380.4 269.6 205.6 591.2 453.9 363.8 835.1 610.3 465.1 1081.6 809.5 640.3 1632.3 1238.7 1033.7 98.1 102.0 71.6 171.4 181.4 152.3 236.3 268.9 224.3 343.6 432.3 331.4 674.6 888.7 550.0 16.7 13.5 9.5 32.1 26.1 18.6 49.2 39.8 31.0 72.6 68.8 56.3 128.0 140.4 109.2 16.7 17.8 14.0 39.3 43.5 36.0 62.6 74.9 67.5 106.4 151.2 108.1 284.8 362.8 244.3 10.7 7.4 6.4 22.7 22.7 19.6 36.5 38.3 30.5 55.4 68.0 48.7 106.8 145.0 88.0 0.9 0.1 0.5 4.7 4.7 3.4 10.2 11.6 8.2 21.5 29.9 18.1 64.9 98.6 56.1 0.0 0.0 0.8 1.2 3.2 3.7 3.1 8.3 9.2 9.4 22.0 24.5 85.0 94.8 73.9 34.4 33.0 14.8 76.8 69.9 53.0 109.4 122.2 102.8 168.9 224.6 180.9 345.1 467.4 367.6
72 0.8 1.8 3.6 6.9 18.2 1.7 4.7 9.4 17.4 42.6 335.8 509.0 649.7 841.6 1197.8 166.1 287.2 383.6 513.2 768.1 122.4 199.0 261.0 365.3 597.5 5.1 9.3 14.4 21.7 43.8 40.5 75.5 114.9 166.3 281.8 19.4 38.1 56.6 79.9 132.0 3.1 8.5 17.1 30.2 64.0 2.2 6.5 14.0 30.4 72.6 44.5 86.0 129.4 194.2 346.9
RI PT
CD4+ late TemRA
0 0.0 0.0 2.7 5.1 13.5 0.0 0.0 1.5 2.9 10.4 316.6 493.3 655.5 886.4 1319.3 182.4 351.1 487.2 695.2 1037.3 57.2 102.8 149.1 214.9 351.2 10.2 27.7 47.7 71.7 147.6 7.9 17.3 29.6 48.4 98.7 5.5 12.2 22.2 32.9 64.3 0.3 2.1 5.8 11.3 27.7 0.0 0.0 0.0 0.0 0.7 18.9 43.1 62.5 96.3 165.1
M AN US C
CD4+ intermediate TemRA
percentile 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th
D
Supplemental Table 2 (continued)
ACCEPTED MANUSCRIPT
B cells
Bnaive
Bmem
AC C
CD27- Bmem
CD27+ Bmem
IgM+ Bmem
Natural effector
IgMonly
72 17.9 33.5 52.1 76.4 139.8 8.9 20.4 33.8 53.4 108.0 4.4 11.9 24.6 52.2 162.6 321.5 480.0 630.0 765.3 1110.0 250.2 389.5 513.6 639.9 933.7 46.0 74.1 98.3 131.2 200.0 20.8 35.8 48.8 65.9 103.8 9.8 17.0 25.2 37.2 65.9 8.7 16.2 25.0 36.8 61.7 6.9 13.0 19.7 30.2 52.1 1.0 2.4 4.1 7.1 15.7
RI PT
Age category (months) 6 14 25 22.3 13.7 5.5 44.5 36.4 28.8 68.3 62.1 54.1 106.0 101.0 89.9 167.0 211.4 169.2 2.6 2.9 1.1 9.8 11.0 6.8 20.8 22.8 17.5 38.9 46.8 34.7 91.3 114.4 79.7 0.0 1.2 1.2 2.8 5.5 5.4 6.3 11.8 12.9 15.6 36.6 33.0 90.4 187.3 151.6 810.0 710.0 520.0 1310.0 1150.0 860.0 1720.0 1490.0 1130.0 2260.0 1950.0 1440.0 3277.5 2590.0 1890.0 733.9 621.0 466.0 1190.7 1001.1 772.7 1609.4 1339.5 1019.3 2104.3 1757.7 1292.0 3030.8 2310.1 1731.0 47.2 59.1 48.8 79.3 98.4 78.3 118.1 142.1 108.2 171.6 191.7 151.5 291.3 305.1 212.6 20.7 27.4 29.4 38.7 52.6 44.2 62.1 72.2 62.8 88.6 98.6 90.5 153.7 157.5 130.0 4.1 6.2 6.3 10.0 18.4 12.0 16.3 28.1 21.1 27.7 42.5 30.9 65.2 75.3 57.1 11.4 12.2 8.9 24.5 26.0 15.5 41.8 41.2 24.7 65.9 68.0 38.1 131.2 133.8 70.2 8.8 9.6 6.9 19.3 20.1 12.3 33.7 33.3 20.4 53.3 55.8 31.3 107.4 113.5 58.3 1.0 0.8 0.9 2.8 3.6 2.4 5.7 6.0 4.0 10.8 10.1 6.5 27.2 25.2 15.6
M AN US C
CD8+ late TemRA
EP
CD8+ intermediate TemRA
0 11.7 29.2 48.0 75.7 123.8 1.4 6.7 11.5 18.8 40.9 0.0 0.0 0.4 1.0 2.1 240.0 467.5 675.0 930.0 1663.5 260.8 425.6 616.3 854.8 1500.9 16.2 31.4 48.0 81.2 168.4 7.7 19.3 29.6 51.7 116.2 0.0 1.2 1.7 3.7 7.0 3.1 6.8 13.0 24.7 67.3 2.7 6.3 12.5 22.3 65.2 0.0 0.0 0.6 1.3 3.6
TE
CD8+ early TemRA
percentile 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th
D
Supplemental Table 2 (continued)
ACCEPTED MANUSCRIPT Age category (months) percentile 0 6 14 25 72 5th 5.8 16.9 16.4 15.0 11.6 25th 11.8 41.3 38.0 30.7 20.4 50th 21.6 70.2 60.8 47.5 28.9 IgA+ Bmem 75th 36.3 117.1 90.9 75.2 40.0 95th 67.9 225.2 151.6 131.3 68.1 5th 4.7 11.3 10.4 10.1 5.0 25th 10.3 31.2 25.3 20.3 10.2 50th 19.0 55.8 43.4 32.4 16.0 CD27-IgA+ 75th 31.9 94.5 67.3 55.6 23.8 95th 62.9 196.7 119.1 101.9 42.6 5th 0.0 3.0 3.3 3.5 4.5 25th 0.9 7.4 9.8 8.1 8.4 50th 1.7 12.0 16.3 13.2 12.2 CD27+IgA+ 75th 3.1 20.1 26.7 21.6 17.3 95th 7.9 42.2 44.2 37.1 27.9 5th 4.2 30.9 33.1 29.7 31.3 25th 8.5 61.0 74.3 61.6 55.7 50th 13.3 87.5 116.8 94.0 77.0 IgG+ Bmem 75th 22.4 143.4 187.6 160.8 102.3 95th 47.9 229.2 318.2 280.1 158.2 5th 3.6 23.3 25.9 23.8 19.2 25th 7.1 48.7 52.6 45.6 36.8 50th 11.2 76.6 87.7 75.4 53.8 CD27-IgG+ 75th 19.2 124.2 151.0 129.2 74.1 95th 42.2 200.3 274.1 224.0 119.3 5th 0.0 3.4 4.4 5.2 7.7 25th 0.8 8.1 14.7 12.6 14.7 50th 1.6 12.3 23.0 19.6 21.1 CD27+IgG+ 75th 3.3 19.2 37.9 33.4 30.1 95th 7.2 35.3 60.8 58.8 51.0 5th 22.0 66.2 60.4 43.1 22.0 25th 38.8 113.7 98.2 74.5 38.4 50th 57.8 157.2 142.8 100.3 55.2 CD21low B cells 75th 91.1 222.4 193.7 134.8 78.1 95th 166.2 351.1 282.9 220.0 124.0 5th 115.0 323.7 290.3 203.9 127.0 25th 220.7 545.5 500.0 351.5 188.8 50th 313.6 757.4 638.0 476.5 251.3 Igλ+ B cells 75th 417.7 975.7 840.0 636.2 317.4 95th 798.4 1486.1 1186.2 847.7 477.2 5th 126.1 456.6 404.8 293.4 179.0 25th 248.8 718.2 635.8 490.5 272.3 50th 352.2 967.0 813.6 634.5 357.0 Igκ+ B cells 75th 493.0 1263.5 1101.9 821.5 439.7 95th 841.4 1857.9 1429.7 1057.4 621.4 The presented reference values are obtained by cross-sectional analyses of all 2,010 data points on the indicated time points
AC C
EP
TE
D
M AN US C
RI PT
Supplemental Table 2 (continued)
ACCEPTED MANUSCRIPT Supplemental Table 3. Overall effect estimates and association with specific age-periods Innate leukocytes Models with 3 knots Determinant
Significance
Granulocytes CD15+ granulocytes Neutrophils Eosinophils Monocytes Classical monocytes Intermediate monocytes
Gender (girl yes/no) Gender (girl yes/no) Low maternal education level Low maternal education level Low maternal education level Seropositivity for CMV at 6yr Siblings >1
** ** ** ** ** ** ****
Determinant
Significance
Antibiotics/Penicillin use in 1st yr
**
NK cells
Effect estimate 359.6 296.4 -63.1 -51.5 -34.3 34.7 -53.2
Standard error 111.6 106.6 24.5 18.1 13.0 12.9 12.1
Effect estimate 6.0
Standard error 2.2
Standard error 120.1 36.1 22.1 20.4 4.6 1.3 1.3 1.7 2.9 3.1 7.2 1.9 37.1 32.4 4.6 5.4 5.0 3.8 3.8 5.3 1.4 21.9 20.3 11.0 10.2
Leukocyte population Non-classical monocytes Lymphocytes Models with 3 knots
M AN US C
Models with 2 knots
Determinant
Significance
T cells TCRαβ+ T cells CD4+ TCRαβ+ T cells
Carrier of H. pylori within 6yrs Gender (girl yes/no) Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Gender (girl yes/no) Low maternal education level Seropositivity for CMV at 6yr Caesarian section Low maternal education level Breastfeeding at 6 months of age Premature rupture of membranes Seropositivity for CMV at 6yr Gender (girl yes/no) Gender (girl yes/no) Gender (girl yes/no) Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Gender (girl yes/no) Seropositivity for EBV at 6y Caesarian section Seropositivity for CMV at 6yr Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Anti-TPO (before 18 weeks of pregnancy) (mU/ml) Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Seropositivity for EBV at 6yr Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Gender (girl yes/no) Seropositivity for CMV at 6yr Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Gender (girl yes/no) Seropositivity for CMV at 6yr
* ** *** * * ** *** ** ** * ** **** ** * ** *** ** ** * * *** *** * **** ****
Effect estimate 303.4 96.9 79.2 51.4 -11.7 3.9 4.9 -4.8 7.9 -6.2 19.5 9.6 107.9 81.3 12.1 20.1 15.5 11.2 9.8 -13.1 5.4 82.3 47.1 104.3 41.2
***
0.02
**** **** **** **** **** **** ** ** **** **** ** *** ****
41.0 31.3 14.1 7.7 9.2 24.8 4.8 -18.1 64.9 8.0 5.5 -13.7 53.3
TCRγδ+ T cells CD4+ TCRγδ+ T cells CD8+ TCRγδ+ T cells Vδ1-Vδ2- TCRαβ- T cells
Vδ1+ T cells CD4+ T cells CD4+ Tnaive CD4+ Tmem CD4+ Tcm CD4+ TemRO
AC C
CD4+ early TemRO CD4+ interm TemRO CD8+ T cells CD8+ Tnaive CD8+ Tmem CD8+ Tcm
EP
TE
Vδ2+ T cells
D
Leukocyte population
CD8+ TCRαβ+ T cells
CD8+ TemRO CD8+ early TemRO CD8+ interm TemRO CD8+ late TemRO CD8+ TemRA CD8+ early TemRA CD8+ interm TemRA CD8+ late TemRA
Significance age-intervals 0-6m 6-14m 14-24m 24-76m ** * **** ns ns ns ns ns ns ns ns ns ns * ** ns * ns ** ns ns ns ns ns **** ns **** ns
RI PT
Leukocyte population
0-14.1m
14.1-70m
70-76m
ns
ns
ns
0-6m
6-14m
14-24m
24-76m
ns ns ns ns ns ns ** ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns * ns
ns ns ns * ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns * ns ns
ns ns * * ns ** ns ns ns ns ns ** ns ns ** ns * ns * ns ns ** ** **** *
* ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns
0.007
****
****
**
****
6.2 5.8 2.6 1.6 1.5 1.9 1.8 6.0 6.5 2.0 1.9 3.6 3.9
ns ns ns ns ns ** ns ** ns ns ns ** **
ns ns ns ns ns ns ** ns ns ns ns * ns
* * * * ** *** ns ns **** ns * ns ****
ns ns ns ns ns ns ns ns ns ns ns ns ns
ACCEPTED MANUSCRIPT Supplemental Table 3 (Lymphocytes; models with 3 knots continued)
Bmem CD27- Bmem CD27+ Bmem IgM+ Bmem Natural effector IgMonly IgA+ Bmem CD27-IgA+ CD27+IgA+ IgG+ Bmem CD27-IgG+ CD27+IgG+ CD21low B cells Igλ+ B cells Igκ+ B cells
Significance
Seropositivity for HSV-1 at 6yr Seropositivity for HSV-1 at 6yr Low maternal education level Seropositivity for EBV at 6yr Low maternal education level Seropositivity for EBV at 6yr Seropositivity for EBV at 6yr Breastfeeding at 6 months of age Carrier of H. pylori within 6yrs Breastfeeding at 6 months of age Carrier of H. pylori within 6yrs Gender (girl yes/no) Carrier of H. pylori within 6yrs Antibiotics/Penicillin use in 1st yr Gender (girl yes/no) Low maternal education level Seropositivity for HSV-1 at 6yr Seropositivity for HSV-1 at 6yr Carrier of H. pylori within 6yrs
** *** * ** * ** * * **** ** **** ** * * ** *** ** ** **
Models with 2 knots
CD4+ TemRA CD4+ early TemRA CD4+ interm TemRA Models with 1 knot Leukocyte population
Significance
Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Premature rupture of membranes Gender (girl yes/no) Premature rupture of membranes
**** *** ** ** **
D
CD4+ late TemRO
Determinant
TE
Leukocyte population
Determinant
Significance
Gender (girl yes/no) *** Seropositivity for CMV at 6yr **** Seropositivity for EBV at 6yr * Vγ9+ T cells Premature rupture of membranes ** Vδ2+Vγ9+ T cells Premature rupture of membranes *** ns= not significant; *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001
AC C
EP
CD4+ late TemRA
Effect estimate -111.2 -108.8 -8.4 -9.3 -4.7 -6.2 -2.7 -5.8 22.0 -4.9 17.8 1.6 3.1 7.1 2.1 -11.7 -53.0 -59.5 82.2
Standard error 35.8 33.0 3.5 3.4 2.0 2.0 1.1 2.3 4.9 1.9 4.1 0.6 1.2 2.8 0.7 3.2 16.6 19.9 30.9
Effect estimate 6.1 2.7 5.0 -0.7 1.9
Standard error 0.7 0.7 1.8 0.2 0.6
Effect estimate -3.0 7.4 1.8 23.1 24.6
Standard error 0.9 0.9 0.9 7.4 6.6
Significance age-intervals 0-6m 6-14m 14-24m 24-76m ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns * ns * ns ns ns ns ns ns * ns ns * **** ns ns ns ns ns ns * **** ns ns ns ns ns ns ns * ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns **
RI PT
B cells Bnaive
Determinant
M AN US C
Leukocyte population
0-14.1m
14.1-70m
70-76m
ns ns ns ns ns
ns ns **** ns ns
* ns ns ns ns
0-25.5m
25.5-76m
ns **** ns * *
ns * ns ns ns
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
A
EP
TE
D
Supplemental Figure 1
all events
granulocytes
granulocytes
monocytes interm
cells/μl
40000 35000 30000 25000 20000 15000 10000 5000 0
classical
12
24
36
48
3500 3000 2500 2000 1500 1000 500 0
60
72
84
96
cells/μl
35000 30000 25000 20000 15000 10000 5000 0
0
12
24
30000
48
60
72
84
10000 5000 0 36
48
60
3000
72
84
0
96
12
24
900 800 700 600 500 400 300 200 100 0
2000 1500 1000 500 0 0
12
24
36
48
60
36
48
60
72
84
age (months)
data points
96
72
84
0
96
12
24
36
48
60
12
450 400 350 300 250 200 150 100 50 0
72
84
96
60
24
36
48
60
72
84
96
72
84
96
non-classical monocytes
0
12
24
36
48
60
age (months)
age (months)
linked data points
48
age (months)
intermediate monocytes
0
36
eosinophils
age (months)
classical monocytes
2500
24
9000 8000 7000 6000 5000 4000 3000 2000 1000 0
15000
24
12
age (months)
20000
12
0
96
neutrophils
25000
age (months)
cells/μl
36
CD3
NK cells
age (months)
CD15+ granulocytes
0
D
3500 3000 2500 2000 1500 1000 500 0
monocytes
age (months)
C
CD14
CD45
CD45
granulocytes
0
eosinophils
lymphocytes
CD45
B
non-classical
CD16
CD16
AC C
CD15
SSC
monocytes
NK cells
CD16/CD56
neutrophils
CD15+ granulocytes
granulocytes
lymphocytes
model
90% CI
72
84
96
A T cells
EP
TCRγδ
CD4+ TCRαβ+ T cells
AC C
CD16/CD56
CD3
TCRγδ+ T cells
CD4+ TCRγδ+ T cells
Vδ1+ T cells
Vδ2+Vγ9+ T cells
Vγ9+ T cells
Vδ1-Vδ2-
TCRαβ
B
CD4
CD4
Vδ1
Vγ9
C
12000
12000
T cells
10000 8000
cells/μl
cells/μl
TCRαβ+ T cells
TCRγδ+ T cells Vδ2+ T cells
CD8+ TCRγδ+ T cells
CD8
CD8+ TCRαβ+ T cells
TCRγδ+ T cells
T cells
TCRγδ+ T cells
TCRαβ+ T cells
CD8
lymphocytes
Vδ2
Supplemental Figure 2
Vδ2
TE
D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
6000 4000 2000 0
0
12
24
36
48
60
72
84
96
5000
8000
4000
6000
3000
4000
2000
2000 0
1000 0
0
12
24
cells/μl cells/μl
1400 1200 1000 800 600 400 200 0 400 350 300 250 200 150 100 50 0
36
48
60
72
84
96
250
TCRγδ+ T cells
CD8+ TCRγδ+ T cells
200
100 50 12
24
36
48
60
72
84
96
Vδ1+ T cells
0
12
24
36
48
0
60
72
84
96
0 900 800 700 600 500 400 300 200 100 0
0
12
24
36
48
60
72
84
96
Vδ2+ T cells
0
12
24
36
48
60
72
84
96
140 120 100 80 60 40 20 0
0 900 800 700 600 500 400 300 200 100 0 0
age (months)
age (months)
data points
12
24
36
48
60
72
84
96
9000 8000 7000 6000 5000 4000 3000 2000 1000 0
CD4+ TCRαβ+ T cells
0
12
24
linked data points
1400 1200 1000 800 600 400 200 0 0 12 24 36 48 60 72 84 96 800 Vγ9+ T cells 700 600 500 400 300 200 100 0 0 12 24 36 48 60 72 84 96
48
60
72
84
96
Vδ1-Vδ2- TCRαβ- T cells
CD4+ TCRγδ+ T cells
12
24
36
48
60
90% CI
72
84
96
Vδ2+Vγ9+ T cells
12
24
36
48
60
age (months)
age (months)
model
36
age (months)
age (months)
150
0
CD8+ TCRαβ+ T cells
age (months)
age (months)
D
6000
TCRαβ+ T cells
10000
72
84
96
RI PT
ACCEPTED MANUSCRIPT
CD4+ T cells Tcm
CD4+ T cells
CD4+ Tmem
TemRA
CD3
B
CD4+
12
24
36
48
60
72
84
cells/μl
9000 8000 7000 6000 5000 4000 3000 2000 1000 0 0 96
early
interm
late
early
late
CD27
CCR7
AC C
9000 8000 7000 6000 5000 4000 3000 2000 1000 0 0
interm
CD4+ TemRA
CD4+ Tnaive
EP
CD8
CD45RO
TemRO
CD4+ TemRO
CD28
TE
T-cells
CD28
A
D
M AN U
SC
Supplemental Figure 3
CD27 2500
CD4+ Tnaive
CD4+ Tmem
2000 1500 1000 500 12
24
36
48
60
72
84
96
0 0
12 24 36 48 60 72 84 96
age (months)
age (months)
age (months)
cells/μl
C
1600 1400 1200 1000 800 600 400 200 0 0
CD4+ Tcm
12
24
36
48
60
72
84
96
1200 1000 800 600 400 200 0 0
1600 CD4+TemRO 1400 1200 1000 800 600 400 200 0 0 12 24 36 48 60 72 84 96
age (months)
CD4+ TemRA
12 24 36 48 60 72 84 96
age (months)
age (months)
cells/μl
D 700 600 500 400 300 200 100 0 0
CD4+ TemRO interm
200 150 100 50
12
24
36
48
60
72
84
96
0 0
12
24
1600 1400 1200 1000 800 600 400 200 0 0
CD4+ TemRA early
12
24
36
48
60
72
36
48
60
72
84
300 250 200 150 100 50 0 0 96
CD4+ TemRO late
12
24
age (months)
age (months)
E
cells/μl
250
CD4+ TemRO early
84
96
age (months) data points
70 60 50 40 30 20 10 0 0
CD4+ TemRA interm
12
24
36
48
60
72
84
200 180 160 140 120 100 80 60 40 20 0 0 96
48
60
72
84
96
84
96
CD4+ TemRA late
12
24
36
48
60
72
age (months)
age (months) linked data points
36
age (months)
model
90% CI
SC
RI PT
ACCEPTED MANUSCRIPT
T-cells
CD8+ Tmem
0
12
24
36
48
60
72
84
96
cells/µl
600 500 400 300 200 100 0 0
CD8+Tcm
2500
CD8+Tnaive
1000 500
00
12
24
36
48
60
72
84
96
00
age (months) 2500
CD8+TemRO
2000
1000 500 60 72 84 96
CD8+Tmem
2000
1500
12 24 36 48
CD27
1500
age (months)
C
late
CD27
3000 2500 2000 1500 1000 500
CD8+
interm
late
EP
6000 5000 4000 3000 2000 1000 0
interm
CCR7
AC C
cells/µl
B
early
early
CD8+ Tnaive
TemRA
CD8+ TemRA
CD28
CD45RO
CD8
Tcm
TE
TemRO
CD8+ T cells
CD3
CD8+ TemRO
CD8+ T cells
CD28
A
D
M AN U
Supplemental Figure 4
00
12 24 36 48 60 72 84 96
age (months) 1200 1000 800 600 400 200 0
CD8+TemRA
0
12 24 36 48 60 72 84 96
age (months)
age (months)
age (months)
12 24 36 48 60 72 84 96
cells/µl
D 1400 1200 1000 800 600 400 200 0
CD8+TemRO early
0
12 24 36 48 60 72 84 96
400 350 300 250 200 150 100 50 00
CD8+TemRO interm
12 24 36 48 60 72 84 96
450 400 350 300 250 200 150 100 50 00
12 24 36 48 60 72 84 96
age (months)
age (months)
age (months)
CD8+TemRO late
cells/µl
E 500 450 400 350 300 250 200 150 100 50 0
CD8+TemRA early
0
12 24 36 48 60 72 84 96
age (months) data points
400 350 300 250 200 150 100 50 0
CD8+TemRA interm
0
12 24 36 48 60 72 84 96
1000 900 800 700 600 500 400 300 200 100 0 0
age (months) linked data points
CD8+TemRA late
12 24 36 48 60 72 84 96
age (months) model
90% CI
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
12 24 36 48 60 72 84 96
age (months)
0
12 24 36 48 60 72 84 96
450 400 350 300 250 200 150 100 50 0 0
age (months)
cells/μl
cells/μl
H
700 600 500 400 300 200 100 0 0 140 120 100 80 60 40 20 0 0
400 350 300 250 200 150 100 50 0 0 12 24 36 48 60 72 84 96 120 IgMonly 100 80 60 40 20 0 0 12 24 36 48 60 72 84 96
Natural effector
age (months)
350 300 250 200 150 100 50 0 0 12 24 36 48 60 72 84 96
age (months)
IgM
IgG+ Bmem
IgA+ Bmem
IgMonly
CD27
800 700 600 500 400 300 200 100 0 0 12 24 36 48 60 72 84 96
D cells/μl 12 24 36 48 60 72 84 96
CD27+ Bmem
12 24 36 48 60 72 84 96
3000 2500 2000 1500 1000 500 0 0
CD27
700 600 500 400 300 200 100 0 0
12 24 36 48 60 72 84 96
450 400 350 300 250 200 150 100 50 0 0 12 24 36 48 60 72 84 96
800 700 600 500 400 300 200 100 0 0 12 24 36 48 60 72 84 96
IgG+ Bmem
12 24 36 48 60 72 84 96
age (months)
age (months)
Cluster 2 Natural effector CD27-IgA+
2 1
CD27+IgG+
12 24 36 48 60 72 84 96
3
CD27-IgG+
36 48 60 72 84 96
Igλ+ B cells
age (months)
IgA+ Bmem
age (months)
I
00
age (months)
IgM+ Bmem
age (months)
IgM
2500 2000 1500 1000 500
Igκ+ B cells
age (months)
age (months)
G
CD27
Bmem
Bnaive
600 500 400 300 200 100 0 0 12 24 12 24 36 48 60 72 84 96 180 CD27+IgA+ 160 140 120 100 80 60 40 20 0 0 12 24 12 24 36 48 60 72 84 96
CD27-IgA+
natural effector
CD27+
CD27
age (months)
CD27+ Bmem CD27+
CD27+
IgM+ Bmem
CD27- Bmem
B cells CD27-
cells/μl
CD21low B cells
5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0 0
B cells CD27-
IgD
EP linked data points model 90% CI
F
800 700 600 500 400 300 200 100 0
C
data points
cells/μl
cells/μl
E
B cells
B cells
Bmem
CD27-
CD21
cells/μl
6000 5000 4000 3000 2000 1000 0 0
Igĸ
CD21low B cells
AC C
cells/μl
B
CD38
Igλ
SSC
Igĸ+ B cells
B cells
CD19
B cells
Bnaive
IgD
B cells
Igλ+ B cells
IgA
B cells
all events
Z-score
A
IgG
TE
D
Supplemental Figure 5
0
Cluster 3 IgMonly CD27-IgG+ CD27+IgA+
-1 -2 -3
Cluster 4 CD27+IgG+
-4 36 48 60 72 84 96
age (months)
-5
0
6
12
18
24
30
36
42
age (months)
48
54
60
66
72
RI PT
ACCEPTED MANUSCRIPT
2
cl4
cl3
3
cl4
cl3
4
4 3 9 6 8 25 52 55 47 43 19 7 37 23 35 40 39 53 56 58 31 41 45 28 44 32 21 33 27 29 59
10
20
30
40 Age (months)
50
6
3
Z-score
2
2
1 0 -1
Z-score
-2 3 2 1 0 -1 -2 -3 -4 -5 2 1
0
6
EP
4 3 9 6
0
12 18 24 30 36 42 48 54 60 66 72
0
1
2
Eosinophils Neutrophils NK cells Classical monocytes
8 25 52 55 47 43 19 7 37 23 35
Non-classical monocytes CD4+ Tcm Natural effector B cells CD27-IgA+ Bnaive CD8+ TemRA early Vδ1+ T cells Intermediate monocytes CD8+ Tcm CD4+ Tnaive CD8+ Tnaive
40 39 53 56 58 31
CD8+ TemRO interm CD8+ TemRO early IgMonly CD27+IgA+ CD27-IgG+ CD4+ TemRA early
41 45 28 44 32 21 33 27 29 59
CD8+ TemRO late CD8+ TemRA late CD4+ TemRO interm CD8+ TemRA interm CD4+ TemRA interm Vδ2+Vγ9+ T cells CD4+ TemRA late CD4+ TemRO early CD4+ TemRO late CD27+IgG+
cluster 1
3
0
6
12 18 24 30 36 42 48 54 60 66 72
4
0 Z-score
-1
12 18 24 30 36 42 48 54 60 66 72
-1 -2 -3 -4 -5
70
TE
1
AC C
Z-score
-2
6 5 4 3 2 1 0 -1 -2
60
z-score
D
0
M AN U
cl2
cl1
1
SC
Supplemental Figure 6
0
6
12 18 24 30 36 42 48 54 60 66 72 Age (months)
cluster 3
ACCEPTED MANUSCRIPT SUPPLEMENTAL MATERIALS AND METHODS Study subjects Included in our study are 1,182 two-generation Dutch children. Peripheral blood was
RI PT
obtained from 220 children at birth, from 376 children at the age of 5-9 months, from 241 children at the age of 13-17 months, from 257 children at the age of 22-30 months and from 916 children at the age of 61-95 months. Of 237 children, detailed immunophenotyping data was available at 3 or more follow-up time points, of 90 children at 4 or more time points. The
M AN US C
complete follow-up at 5 time points was available for 12 children.
Sample preparation
Approximately 4ml of blood was collected in EDTA-anticoagulant tubes by venous puncture. Blood was kept at room temperature and processed within 24 hours after sampling. 50µl Full blood was mixed with BD Trucount beads (BD Biosciences; San Jose, CA) and stained for CD3, CD16, CD56, CD19 and CD45 in a diagnostic lyse-no-wash approach to obtain absolute cell counts. Subsequently, red blood cells were lysed by adding NH4Cl. Absolute
D
numbers of leukocytes, NK cells, T cells and B cells were obtained without a wash step on a BD FACSCalibur (BD Biosciences; San Jose, CA).
TE
2ml Full blood was subjected to bulk lysis of red blood cells by 10 min incubation with NH4Cl. Following wash steps, the remaining white blood cells were divided over 10
EP
tubes for 6-color flowcytometric analyses with the antibodies listed in Supplemental Table 1. Following 10 min incubation at room temperature, cells were washed and immunophenotypic
AC C
measurements were performed on a 3-laser BD LSRII (BD Biosciences; San Jose, CA). The 488nm laser was used to excite the FITC (530/30nm filter), PE (575/26nm), PerCP-Cy5.5 (695/40nm) and PE-Cy7 (780/60nm) fluorophores, and the 633nm laser for APC (660/20nm) and APC-Cy7 (780/60nm). PMT voltages were set using 8-peak rainbow beads based on target values for the 7th peak as established within the EuroFlow consortium.27 Absolute cells per µl blood of leukocyte subpopulations were recalculated using the total leukocyte, NKcell, T-cell and B-cell numbers obtained from the Trucount analysis.
Statistical modeling Our dataset included 62 leukocyte populations and 26 determinants for 1,182 children, with for each leukocyte population (defined as population “i”), a total of 2,010 data points.
Linear mixed effect modeling of leukocyte kinetics 1
ACCEPTED MANUSCRIPT For each leukocyte population “i” in our dataset, the following 4 linear mixed effect models were tested:
lmer(dataset[,i] ~ ns(Age,1)+(1|Id),REML=FALSE,data=data)
lmR_1knots<-
RI PT
lmR_0knots<-
M AN US C
lmer(dataset[,i] ~ ns(Age,2, knots = c(25.5))+(1|Id),REML=FALSE,data=dataset)
lmR_2knots<-
lmer(dataset[,i] ~ ns(Age,3, knots = c(14.1 , 70))+(1|Id),REML=FALSE,data=dataset)
lmR_3knots<-
lmer(dataset[,i] ~ ns(Age,4, knots = c(6, 14, 24))+(1|Id),REML=FALSE,data=dataset)
# ns(Age,...) defines the natural spline of the age of the children, with indicated knots;
TE
longitudinal measurements.
D
# (1|Id) defines the child’s Id-number as a random effect in the model, to include
EP
The optimal model was selected by likelyhood ratio test:
anov[1]<-ifelse(anova(lmR_0knots,lmR_1knots)[[8]][2]<0.05,2,0)
AC C
anov[2]<-ifelse(anova(lmR_1knots,lmR_2knots)[[8]][2]<0.05,3,0) anov[3]<-ifelse(anova(lmR_2knots,lmR_3knots)[[8]][2]<0.05,4,0) selected.lmR<-max(anov)
Univariate analyses of the effect of nongenetic determinants on leukocyte kinetics Each determinant “j” in our dataset was included independently as a fixed effect to the optimal linear mixed effect model of individual leukocyte populations (below an example for a leukocyte population modelled by model lmR_3knots) lmer(dataset[,i] ~ ns(Age,4, knots = c(6,14,24))+determinant[,j]+(1|Id),REML=FALSE,data=dataset)
Determinants with an effect of p<0.0125 were defined as significant. 2
ACCEPTED MANUSCRIPT Multivariable analyses of the effect of nongenetic determinants on leukocyte kinetics All determinants with a significant univariate effect on a leukocyte population (up to three determinants per leukocyte population), were combined and included as fixed effects to the
RI PT
optimal linear mixed effect model of individual leukocyte populations to correct for confounding effects (below an example for a leukocyte population modelled by model lmR_3knots)
M AN US C
lmer(dataset[,i] ~ ns(Age,4, knots = c(6,14,24))+significant determinant 1+ significant determinant 2+ etc +(1|Id),REML=FALSE,data=dataset)
Determinants with an effect of p<0.05 were defined as significant after multivariable correction.
Define whether a determinant has an age-associated affect on a leukocyte population For determinants that still affected leukocyte kinetics after multivariable correction, we
D
subsequently defined whether this effect was significantly stronger within a selective ageperiod (either between the age of 0months – knot1, knot1 – knot2 , knot2 – knot3 or knot3 –
TE
76 months). To test this, the models for multivariable analyses were adjusted, testing
EP
individual determinants in relation to the age in the children:
e.g. Test the age-associated effect of determinant 1: lmer(dataset[,i] ~ ns(Age,4, knots = c(6,14,24))* significant determinant 1 + significant
AC C
determinant 2+ etc +(1|Id),REML=FALSE,data=dataset)
# “*” indicates a test of the natural spline of the age of the children in relation to determinant 1 # the other determinants significantly affecting this leukocyte population were still included as fixed effects to correct for potential confounding effects
3
ACCEPTED MANUSCRIPT
M AN US C
RI PT
SUPPLEMENTAL FIGURES: FIGURE LEGENDS Supplemental Figure 1. Dynamics of innate leukocyte subsets in children between birth and 6 years of age. A) Gating strategy used to identify innate leukocyte subsets. Plots depict data of a representative 61-95 months-old child. B-D) Absolute numbers and modelled dynamics of granulocytes, monocytes and NK cells (B), CD15+, neutrophilic and eosinophilic granulocytes (C), and classical, intermediate and non-classical monocyte subsets (D), from birth until 6 years of age. Linear mixed effect models were generated for each population (solid black line) and represented with the 90% confidence interval (CI) of the model (dashed black lines). For clarity of the graphs, gray lines connect only consecutive time points within one individual; i.e. 0-6m, 6-14m, 14-25m or 25-76m.
D
Supplemental Figure 2. Dynamics of TCRαβ+ T-cell and TCRγδ+ T-cell subsets in children between birth and 6 years of age. A) Gating strategy used to identify TCRαβ+ T-cell and TCRγδ+ T-cell subsets. Plots depict data of a representative 61-95 months-old child. B-D) Absolute numbers and modelled dynamics of T cells (B), TCRαβ+ T-cell populations (C), and TCRγδ+ T-cell populations (D), from birth until 6 years of age. Linear mixed effect models were generated for each population (solid black line) and represented with the 90% confidence interval (CI) of the model (dashed black lines). For clarity of the graphs, gray lines connect only consecutive time points within one individual; i.e. 0-6m, 6-14m, 14-25m or 25-76m.
AC C
EP
TE
Supplemental Figure 3. Dynamics of CD4+ T-cell subsets in children between birth and 6 years of age. A) Gating strategy used to identify CD4+ T-cell subsets. Plots depict data of a representative 61-95 months-old child. B-E) Absolute numbers and modelled dynamics of total, naive and memory CD4+ T cells (B); CD4+ central memory T cells (Tcm), CD45RO+ effector memory (TemRO) and CD45RO- effector memory (TemRA) T-cell subsets (C); early, intermediate or late differentiated TemRO subsets (D); or early, intermediate or late differentiated TemRA subsets (E), between birth and 6 years of age. Linear mixed effect models were generated for each population (solid black line) and represented with the 90% confidence interval (CI) of the model (dashed black lines). For clarity of the graphs, gray lines connect only consecutive time points within one individual; i.e. 0-6m, 6-14m, 14-25m or 25-76m. Supplemental Figure 4. Dynamics of CD8+ T-cell subsets in children between birth and 6 years of age. A) Gating strategy used to identify CD8+ T-cell subsets. Plots depict data of a representative 61-95 months-old child. B-E) Absolute numbers and modelled dynamics of total, naive and memory CD8+ T cells (B); CD8+ central memory T cells (Tcm), CD45RO+ effector memory (TemRO) and CD45RO- effector memory (TemRA) T cell subsets (C); early, intermediate or late differentiated TemRO subsets (D); or early, intermediate or late differentiated TemRA subsets (E), between birth and 6 years of age. Linear mixed effect models were generated for each population (solid black line) and represented with the 90% confidence interval (CI) of the model (dashed black lines). For clarity of the graphs, gray lines connect only consecutive time points within one individual; i.e. 0-6m, 6-14m, 14-25m or 25-76m. Supplemental Figure 5. Dynamics of B-cell subsets in children between birth and 6 years of age. 4
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
RI PT
A) Gating strategy used to identify B-cell subsets. Plots depict data of a representative 61-95 months-old child. B-I)Absolute numbers and modelled dynamics of total B cells (B); naive and total memory B cells (C); total Igκ+ and Igλ+ B-cells populations (D); and CD21low B cells (E); CD27- and CD27+ memory B cells (F); total IgM+, IgA+ or IgG+ memory B cells (G); IgM+, IgA+ or IgG+ memory B-cell subsets (H) between birth and 6 years of age. Linear mixed effect models were generated for each population (solid black line) and represented with the 90% confidence interval (CI) of the model (dashed black lines). For clarity of the graphs, gray lines connect only consecutive time points within one individual; i.e. 0-6m, 614m, 14-25m or 25-76m. I) Overlay of the normalized (zero mean; unit SD) linear mixed effect models of the memory B-cell populations as in panel H.
5
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
RI PT
Supplemental Figure 6. Hierarchical clustering of the dynamics of 31 non-overlapping leukocyte subsets in early childhood. A) Ward’s hierarchical clustering was performed as in Figure 2, including only the 31 nonoverlapping leukocyte subsets. Indicated in gray squares in front of the heat map is the cluster that each population was assigned to in the analysis of the total 62 leukocyte subsets in Figure 2. B) Average patterns +/- 1 standard deviation of the subsets in each of the 4 major clusters. Indicated in black squares are 3 subsets that were assigned to different clusters than in Figure 2.
6
S0091-6749(16)31379-3
DOI:
10.1016/j.jaci.2016.10.023
Reference:
YMAI 12477
To appear in:
Journal of Allergy and Clinical Immunology
Received Date: 17 April 2016 Revised Date:
29 August 2016
Accepted Date: 5 October 2016
Please cite this article as: van den Heuvel D, Jansen MAE, Nasserinejad K, Dik WA, van Lochem EG, Bakker-Jonges LE, Bouallouch-Charif H, Jaddoe VWV, Hooijkaas H, van Dongen JJM, Moll HA, van Zelm MC, Effects of nongenetic factors on immune cell dynamics in early childhood: The Generation R Study, Journal of Allergy and Clinical Immunology (2016), doi: 10.1016/j.jaci.2016.10.023. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
TE
D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
Immunophenotyping of Children Aged 0 - 6 Years Lymphocytes NK cells
Granulo
T cells
AC C
Mono
Lympho
Z- score
CD3
Naive
Naive
Mem
CD8+ T cells
Mem
CD4+ T cells
Mem
B cells Naive
NK cells
Monocytes
Granulocytes
Effects of Nongenetic Determinants on Leukocyte Kinetics 62 Leukocyte Populations
26 Nongenetic Determinants
Birth Characteristics (n=6) Bacterial/Viral Exposure (n=11) Distinct effects of various determinants in the indicated category
Z- score
Maternal Immune-mediated Diseases (n=3)
Decrease, or
Innate Leukocytes (n=8)
2
Naive Lymphocytes (n=24)
3
Memory Lymphocytes (n=18)
2.00
4
Memory Lymphocytes (n=12)
1.00
Maternal Life Style (n=6)
Increase
1
Z- score
CD45
7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00 -1.00 -2.00 3.00 2.50 2.00 1.50 1.00 0.50 0.00 -0.50 -1.00 -1.50 -2.00 3.00 2.00 1.00 0.00 -1.00 -2.00 -3.00 -4.00 -5.00
Z- score
CD16/CD56
EP
All Events
SSC
Kinetics Profiles of 62 Leukocyte Subsets
0.00 -1.00 -2.00 -3.00 -4.00 -5.00
0
6
12 18 24 30 36 42 48 54 60 66 72 Age (months)
ACCEPTED MANUSCRIPT Effects of nongenetic factors on immune cell dynamics in early childhood: The
Running title: Determinants of leukocyte subset kinetics in children
RI PT
Generation R Study
Diana van den Heuvel, PhD,a Michelle A.E. Jansen, MD,b,c Kazem Nasserinejad, MSc,d Willem A. Dik, PhD,a Ellen G. van Lochem, PhD,a# Liesbeth E. Bakker-Jonges, PhD,a‡
M AN US C
Halima Bouallouch-Charif, BSc,a Vincent W.V. Jaddoe, MD, PhD,b,c,e Herbert Hooijkaas, PhD,a Jacques J.M. van Dongen, MD, PhD,a* Henriëtte A. Moll, MD, PhD,c* Menno C. van Zelm, PhDa,f* a
Department of Immunology, Erasmus MC, University Medical Center, Rotterdam, the
Netherlands;
b
The Generation R Study Group, Erasmus MC, University Medical Center,
Rotterdam, the Netherlands; the Netherlands;
d
c
Department of Pediatrics, Erasmus MC-Sophia, Rotterdam,
Department of Biostatistics, Erasmus MC, University Medical Center, e
Department of Epidemiology, Erasmus MC, University
D
Rotterdam, the Netherlands;
Medical Center, Rotterdam, the Netherlands;
f
Department of Immunology and Pathology,
TE
Central Clinical School, Monash University, Melbourne, Victoria, Australia;
#
Present
address: Department of Microbiology and Immunology, Rijnstate Hospital, Arnhem, the
EP
Netherlands; ‡ Present address: Department of Medical Laboratory, Reinier de Graaf Groep, Delft, The Netherlands;
AC C
* shared senior authorship
Correspondence:
Menno C. van Zelm, PhD
Department of Immunology and Pathology, Central Clinical School, Monash University, 89 Commercial Road, Melbourne, Victoria 3004, Australia Phone: +61 39903 0834 Email: [email protected]
This work was supported by an Erasmus MC Fellowship to M.C.v.Z
1
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
anti-TPO, anti-thyroid peroxidase IgG CI, confidence interval class, classical monocytes CMV, cytomegalovirus EBV, Epstein Barr virus eos, eosinophils H pylori, Helicobacter pylori HSV-1, herpes simplex virus 1 inter, intermediate monocytes mem, memory n-class, non-classical monocytes neu, neutrophils PROMM, premature rupture of membranes RTI, respiratory tract infection Tcm, central memory T cells Tem, effector memory T cells TG2A, anti-tissue transglutaminase IgA antibody VZV, varicella zoster virus
RI PT
Abbreviations:
2
ACCEPTED MANUSCRIPT ABSTRACT
2
Background. Numbers of blood leukocyte subsets are highly dynamic in childhood and
3
differ greatly between individuals. Inter-individual variation is only partly accounted for by
4
genetic factors.
5
Objective. Determine which nongenetic factors affect the dynamics of innate leukocytes, and
6
naive and memory lymphocyte subsets.
7
Methods. We performed six-color flow cytometry and linear mixed effect modeling to define
8
the dynamics of 62 leukocyte subsets from birth to 6 years of age in 1,182 children with one
9
to five measurements per individual. Subsequently, we defined the impact of prenatal
10
maternal lifestyle-related or immune-mediated determinants, birth characteristics and
11
bacterial/viral exposure-related determinants on leukocyte subset dynamics.
12
Results. Functionally similar leukocyte populations were grouped by unbiased hierarchical
13
clustering of patterns of age-related leukocyte dynamics. Innate leukocyte numbers were high
14
at birth and were predominantly affected by maternal low education level. Naive
15
lymphocytes peaked around 1 year, while most memory lymphocyte subsets more gradually
16
increased during the first 4 years of life. Dynamics of CD4+ T cells were predominantly
17
associated with gender, birth characteristics, and persistent infections with cytomegalovirus
18
(CMV) or Epstein Barr virus (EBV). CD8+ T cells were predominantly associated with CMV
19
and EBV infections, and TCRγδ+ T cells with premature rupture of membranes and CMV
20
infection. B-cell subsets were predominantly associated with gender, breastfeeding and
21
Helicobacter pylori carriership.
22
Conclusions. Our study identifies specific dynamic patterns of leukocyte subset numbers, as
23
well as nongenetic determinants that affect these patterns, thereby providing new insights into
24
the shaping of the childhood immune system.
AC C
EP
TE
D
M AN US C
RI PT
1
25
3
ACCEPTED MANUSCRIPT 26 27
KEY MESSAGES •
Absolute numbers of leukocyte populations are differ between individuals, but functionally related leukocyte populations follow similar age-related leukocyte dynamics
29
between birth and 6 years of age.
30
•
RI PT
28
We identified nongenetic factors that are associated with the dynamics of innate leukocyte, B-cell and CD4+ and CD8+ T-cell lineages (as well as specific leukocyte
32
subsets) and that underlie, at least in part, the human immunological diversity.
M AN US C
31
33
CAPSULE SUMMARY
35
Human inter-individual immunological diversity can only partly be explained by genetic
36
factors, indicating an important contribution of nongenetic factors. We here analyzed
37
leukocyte dynamics in 1,182 children of the Generation R Study and determined the effect of
38
26 nongenetic factors on inter-individual immunological diversity.
TE
D
34
39
KEY WORDS
41
Longitudinal
42
leukocytes, CD4+ T cells, CD8+ T cells, B cells, TCRγδ+ T cells, nongenetic determinants.
leukocyte
dynamics,
inter-individual
AC C
43
EP
40
4
immunological
variance,
innate
ACCEPTED MANUSCRIPT INTRODUCTION
45
The human immune system shows high diversity in cellular composition and functional
46
responses between individuals. Blood leukocytes in young children are highly dynamic in
47
numbers and composition.1-6 Innate cell numbers, such as neutrophils and NK cells, are
48
higher in neonates than in children or adults,1, 5, 7 and already display dynamic changes within
49
the first few days after birth.1, 5, 6 B and T cells are mostly naive in infants, while protective
50
immunity is gradually built up in the form of increasing numbers of memory B and T cells
51
during the first 5 years of life after which these numbers stabilize.2-4, 8-11 Recent studies have
52
found that 25-50% of inter-individual variation in cellular composition and functional
53
responses were accounted for by genetic factors,12, 13 indicating that nongenetic determinants
54
underlie a large part of immune trait variance,14 which will potentially have long-term effects
55
and underlie part of the immune trait variation in adults.
D
M AN US C
RI PT
44
To date, correlations of patterns between different immune cells are incompletely
57
studied, due to either the restricted numbers of analyzed subsets, the short-term follow-up, or
58
the lack of longitudinal analyses correcting for intrapersonal correlations between repeated
59
measures. Furthermore, the nongenetic determinants that drive the kinetics of each type of
60
immune cell remain less well studied. Likely, these involve various environmental
61
determinants, such as maternal life style, maternal immune-mediated diseases, birth
62
characteristics, and bacterial and viral exposure in childhood.15-24 Studying the impact of
63
these nongenetic factors on the dynamics of childhood immune development requires large
64
cohorts of healthy young children, with multiple measurements per individual.
AC C
EP
TE
56
65
We investigated which nongenetic factors, related to the prenatal maternal lifestyle,
66
prenatal maternal immune-mediated diseases, birth characteristics or bacterial/viral exposure-
67
related characteristics, influence the dynamics of blood leukocyte populations from birth until
5
ACCEPTED MANUSCRIPT 68
6 years of age. This concerned a total of 62 leukocyte populations, including innate leukocyte
69
subsets, naive and memory B-cell and T-cell subsets.
AC C
EP
TE
D
M AN US C
RI PT
70
6
ACCEPTED MANUSCRIPT METHODS
72
Study subjects
73
This study was embedded in the Generation R Study, a prospective population-based cohort
74
study from fetal life until young adulthood.25, 26 The current study focused on a subgroup of
75
1,182 two-generation Dutch children, born between August 2003 and August 2006.
76
Peripheral blood was obtained at birth, and median age of 6 months, 14 months, 25 months
77
and 72 months. Detailed immunophenotyping was performed at 1-5 time points per child,
78
resulting in a total of 2,010 data points (details in Suppl Methods). Written informed consent
79
was obtained from parents, according to the Medical Ethics Committee guidelines of
80
Erasmus MC.
M AN US C
RI PT
71
81
Immunophenotyping
83
Absolute numbers of leukocytes, NK cells, T cells and B cells were obtained with a
84
diagnostic lyse-no-wash protocol, using a BD FACSCalibur. Six-color flow cytometry was
85
performed to quantify 62 well-defined leukocyte populations (Table 1), which included NK
86
cell, monocyte, granulocyte, naive and memory B-cell, naive and memory TCRαβ+ T-cell
87
and TCRγδ+ T-cell subsets (Supplemental Table 1). Flowcytometric analyses were performed
88
on an LSRII (BD Biosciences) using standardized measurement settings (details in Suppl
89
Methods).27
AC C
EP
TE
D
82
90 91
Determinants
92
Information on 22 dichotomized and 4 continuous determinants was obtained (Table 2).25, 26
93
Six determinants were related to prenatal maternal life style, and were evaluated using
94
questionnaires in the first, second and third trimester of pregnancy:26 maternal age and body
95
mass index before pregnancy, education, net household income, and smoking or alcohol use
7
ACCEPTED MANUSCRIPT during pregnancy. Three determinants were related to prenatal maternal immune-mediated
97
diseases: data on maternal atopy was obtained by maternal reported questionnaires during
98
pregnancy,26 serum anti-thyroid peroxidase IgG (anti-TPO) and anti-tissue transglutaminase
99
IgA antibody (TG2A) levels were measured in the second trimester of pregnancy (mean ±
100
SD; 13.4 ± 2.0 weeks; and 20.6 ± 1.2 weeks, respectively).28, 29 Information on 6 birth-related
101
determinants, i.e. gender, gestational age (preterm birth <37w), birth weight (low birth weight
102
<2,500g), premature ruptures of membranes, mode of delivery and birth season was obtained
103
by obstetric records assessed in midwife practices and hospitals.25, 26 Eleven bacterial or viral
104
infection-related determinants were included: breastfeeding and breastfeeding duration,
105
having siblings, antibiotics usage and presence of upper or lower respiratory tract infection in
106
the first year of life (obtained using questionnaires at the child’s age of 2, 3, 6 and 12
107
months),26 Helicobacter pylori (H. pylori) carriership at 6y of age, and seropositivity (IgG)
108
for cytomegalovirus (CMV), Epstein Barr virus (EBV), herpes simplex virus 1 (HSV-1) and
109
varicella zoster virus (VZV) at the age of 6 years.22, 30, 31
110
EP
TE
D
M AN US C
RI PT
96
Statistical modeling
112
To model leukocyte dynamics between birth and the age of 6 years, linear mixed effect
113
analyses were performed. By including random-effects in the model, this approach enabled
114
modeling of cross-sectional data, with further improvement of the accuracy by incorporating
115
longitudinal follow-up data from individual children. To capture the trend in the data more
116
precisely, the age of the children was included as a natural spline with different knots (0-3
117
knots) into the models. Basically, the number of knots is inversely related to the smoothness
118
of the curve. Positions of the knots were defined as the 50th percentile (25.5 months) in the 1-
119
knot model, and the 33rd and 66th percentiles (14.1 and 70 months) in the 2-knots model. The
AC C
111
8
ACCEPTED MANUSCRIPT 120
knots in the 3-knot model were defined manually at 6, 14 and 24 months, focusing around the
121
time points of data measurement. Model selection was performed by likelihood ratio test. Next, for each leukocyte population the effect of all 26 determinants on the models was
123
assessed by first adding each determinant univariably into the model and analyzing fixed
124
effect estimates. To correct for potential multiple testing errors, a correction for the four
125
groups of determinants was performed, and consequently only determinants with an effect of
126
p<0.0125 were defined as significant. Subsequently, for each leukocyte population, the
127
determinants with a significant effect (up to 3 determinants were significant per model) were
128
combined in a multivariable model to correct for possible confounding effects. Last, to test
129
whether changes in leukocyte dynamics over time (from birth until 6 years), induced by the
130
determinant of interest was most noticeable at a specific age interval the individual
131
determinants were tested in relation to the age of the children in a multivariable model. The
132
p-value for significance for the multivariable analyses was p<0.05. Statistical analyses were
133
performed in R (version R-3.2.1; details in Supplemental Material and Methods).32
134
EP
TE
D
M AN US C
RI PT
122
Clustering analyses
136
To cluster patterns of leukocyte subsets kinetics, the modelled data of each population was
137
normalized into zero mean and unit standard deviation (z-score), using the following
138
calculation:
139
z-score =
140
The normalized leukocyte models were subsequently clustered using agglomerative (‘bottom-
141
up’) Ward’s hierarchical clustering, at each step clustering two clusters with minimum
142
between-cluster distance, using the Euclidean distance measure.
AC C
135
9
ACCEPTED MANUSCRIPT RESULTS
144
Linear mixed effect modeling of leukocyte-subset cell numbers versus the child’s age
145
To study the dynamics of blood immune cells in young children, we quantified cell numbers
146
of 11 leukocyte subsets in 1,182 children between birth and 76 months (6y) of age. Statistical
147
modeling showed that the numbers of four innate leukocyte subsets, i.e. monocytes,
148
neutrophils, eosinophils and NK cells, were high at birth, quickly declined within the first 6
149
months of age and subsequently remained stable (Figure 1A). Naive B-cell and T-cell
150
numbers strongly increased after birth, peaked between 6-14 months of age and subsequently
151
decreased to stable levels around the age of 2-6 years (Figure 1B). Memory B-cell and T-cell
152
numbers slowly increased within the first 6-14 months of life, after which numbers declined
153
marginally and stabilized from ~3 years onwards (Figure 1B). Total TCRγδ+ T-cell numbers
154
increased until 6 months and remained quite stable at these levels (Figure 1C).
D
M AN US C
RI PT
143
TE
155
Distinct dynamics of innate leukocyte, and naive and memory B and T cells
157
To evaluate whether functionally related immune populations would follow similar dynamics
158
with increasing age, we extended our analysis in the same children to 62 leukocyte subsets,
159
based on recent insights into naive and memory B-cell and T-cell subsets (Table 1 and Suppl
160
Figures 1-5; reference values per age category are presented in Suppl Table 2).24, 33
AC C
EP
156
161
Hierarchical clustering of all 62 leukocyte populations resulted in 4 distinct clusters of
162
leukocyte dynamics (Figure 2A). Cell numbers of all populations in the first cluster were high
163
at birth, followed by a sharp decrease within the first 6 months of life, after which they
164
stabilized (Figure 2B). This cluster exclusively contained innate leukocyte subsets. All
165
defined innate leukocyte subsets clustered within cluster 1, except for the intermediate
166
monocytes.
10
ACCEPTED MANUSCRIPT The populations in cluster 2 strongly increased shortly after birth, peaked before 14
168
months of age, followed by a long-term gradual decrease. Cluster 2 included the 3 major
169
naive lymphocyte subsets (CD4+ Tnaive, CD8+ Tnaive and Bnaive), as well as 5 memory
170
populations (CD4+ central memory T cells (Tcm), CD8+ Tcm cells, early differentiated CD8+
171
CD45RO- effector memory T cells (CD8+ early TemRA), CD27-IgA+ memory B cells and
172
natural effector memory B cells). Also the total B-cell, CD8+ T-cells, CD4+ T-cells and total
173
T-cell subsets clustered in cluster 2, as well as the intermediate monocytes and the Vδ1+ T
174
cells.
M AN US C
RI PT
167
The populations in clusters 3 and 4 gradually increased in cell number and peaked either
176
at 14 months (cluster 3) or after ~4 years (cluster 4). Clusters 3 and 4 contained only memory
177
B-cell and T-cell subsets, and the TCRγδ+ T-cell subsets. Cluster 3 contained the total
178
memory lymphocyte populations Bmem, CD4+ Tmem and CD8+ Tmem. In particular, these
179
memory populations included the early, intermediate and late CD8+ TemRO subpopulations,
180
late CD8+ TemRA cells, early CD4+ TemRA cells, and the IgMonly, CD27+IgA+ and CD27-
181
IgG+ memory B-cell subsets. Cluster 4 contained total TCRγδ+ T cells and the Vδ2+ and the
182
Vγ9+ T-cell subsets, as well as the early, intermediate and late CD4+ TemRO subpopulations,
183
late CD4+ TemRA cells, intermediate CD8+ TemRA cells, and the CD27+IgG+ memory B
184
cells.
AC C
EP
TE
D
175
185
To test the effect of including populations with overlapping population definitions on
186
the clustering, an additional clustering was performed on a selection of 31 non-overlapping
187
populations (Table 1). The resulting clusters showed similar dynamics (Figure 2B and
188
Supplemental Figure 6B) with only 3 populations (non-classical monocytes, CD8+ late
189
TemRO and CD8+ late TemRA) being assigned to a different cluster, indicating that the
190
overlapping populations did not overtly skew our clustering approach. Thus, functionally
11
ACCEPTED MANUSCRIPT 191
related populations, as well as populations with phenotypic overlapping definitions displayed
192
similar dynamics in early childhood.
RI PT
193
Associations between nongenetic factors and leukocyte subset dynamics in the four
195
clusters
196
The memory B- and T-cell subset showed large inter-individual variation in dynamics of cell
197
numbers (Figure 1). To study the effects of external factors, we here analyzed 6 maternal life
198
style- and 3 maternal immune-related determinants, 6 birth characteristics, and 11
199
bacterial/viral exposure-related determinants (Table 2). Twelve of these 26 determinants
200
showed a significant association with the dynamics of one or more of the 62 leukocyte
201
populations after multivariable correction (Table 2). Bacterial/viral exposure-related
202
determinants were more frequently found to affect cell numbers in clusters 2, 3 and 4 than in
203
cluster 1 (Figure 3A-B).
TE
D
M AN US C
194
In cluster 1, a low maternal education level was associated with a reduction in the
205
patterns of eosinophils and classical monocytes (Figure 4). Female gender was associated
206
with a significant increase in the pattern of neutrophils, and the phenotypically overlapping
207
CD15+ granulocyte population. Having more than 1 sibling was associated with a reduction,
208
and CMV infection with an increase, in NK cells. Antibiotics usage in the first year of life
209
was associated with an increase in non-classical monocytes. Overall effect estimates and
210
associations of effects with specific age-intervals are presented in Suppl Table 3.
AC C
EP
204
211
In cluster 2, none of the determinants included in our study affected the patterns of more
212
than 25% of the populations (Figure 3C-D). However, female gender was associated with an
213
increase in the patterns of CD4+ naive and Tcm cells, and consequently in the phenotypically
214
overlapping CD4+TCRαβ+ and total CD4+ T cells. Whereas breastfeeding for more than 6
215
months was associated with a reduction in the pattern of total IgA+ and CD27-IgA+ memory
12
ACCEPTED MANUSCRIPT B cells, H. pylori carriership was associated with an increase in these populations.
217
Furthermore, H. pylori carriership was associated with an increase in total Igκ+ B cells and
218
TCRαβ+ T cells. CMV and EBV were both associated with an increase in the strongly related
219
total and TCRαβ+ CD8+ T cells, and CMV with an increase in Vδ1+ T cells. HSV-1
220
seropositivity was associated with a decrease in the large population of naive B cells, and
221
consequently with total B cells and Igκ+ and Igλ+ B-cell subsets.
M AN US C
RI PT
216
In cluster 3, CMV and EBV seropositivity were significantly associated with changes in
223
dynamics of >40% of subsets (Figure 3C; Suppl Table 3). Both viruses were associated with
224
an increase in CD8+ intermediate and late TemRO cells, and the phenotypically-related total
225
CD8+ TemRO and total CD8+ Tmem populations. In addition, CMV was associated with an
226
increase in CD8+TCRγδ+ T cells, and CD8+ late TemRA cells, and the phenotypically-related
227
total CD8+ TemRA population. EBV was associated with an increase in CD8+ early TemRO
228
cells, and a decrease in total Bmem, and the CD27- and CD27+ Bmem subsets. The effects on
229
CD8+ late TemRO and TemRA cells were still present in the analysis of 31 non-overlapping
230
population, even though these were now included in cluster 4 (Figure 3D).
EP
TE
D
222
In cluster 4, gender, premature rupture of membranes, CMV and EBV were significantly
232
associated with changes in dynamics of 40% of subsets (Figure 3C-D; Suppl Table 3).
233
Female gender was associated with an increase in the patterns of CD27+IgG+ memory B cells,
234
and CD4+ early TemRO cells, and with a reduction in total TCRγδ+ T cells and CD4+
235
intermediate and late TemRA populations. Premature rupture of membranes was associated
236
with an increase in CD4+ late TemRO, CD4+ intermediate TemRA and Vδ2+ and Vγ9+ T-cell
237
subsets. CMV and EBV seropositivity were both associated with an increase in total CD4+
238
TemRO and CD4+ late TemRO and TemRA cells, as well as CD8+ intermediate TemRA
239
cells. In addition, CMV was associated with increased CD4+ intermediate TemRO cells, and
240
EBV with an increase in CD4+ early TemRO cells.
AC C
231
13
ACCEPTED MANUSCRIPT Altogether, we identified multiple determinants that had a significant effect on the
242
dynamics of leukocyte subset numbers within one or more of the four distinct age-related
243
patterns. Still, considerable variation of effects could be observed within individual clusters.
RI PT
241
244
Effects of nongenetic factors on leukocyte dynamics within distinct leukocyte lineages
246
As clusters 2, 3 and 4 each contained various B-cell and T-cell populations, each with distinct
247
humoral or cellular functions, we next studied the effects of the 26 determinants on the
248
patterns of individual leukocyte lineages.
M AN US C
245
Within the B-cell lineage, low maternal educational level was associated with a reduction
250
in the patterns of total memory B cells (Figure 4). Female gender was associated with an
251
increase in the patterns of CD27+IgA+ and CD27+IgG+ memory B cells, though these effects
252
were not reflected in the total memory B-cell populations. Breastfeeding for more than 6
253
months was associated with a selective reduction in the pattern of CD27-IgA+ memory B
254
cells, antibiotics usage in the first year of life was associated with a selective increase in the
255
pattern of CD27-IgG+ memory B cells, and H. pylori carriership was associated with an
256
increase in the patterns of both CD27-IgA+ and CD27+IgA+ memory B cells. EBV infection
257
was associated with a reduction in the pattern of total memory B cells, though not with a
258
specific memory B-cell subset, HSV-1 infection was associated with a selective reduction in
259
the pattern of naive B cells.
AC C
EP
TE
D
249
260
Within the CD4+ and CD8+ T-cell lineage increased levels of serum anti-TPO was
261
associated with an increase in CD8+ Tcm cells. Female gender was associated with an
262
increase in the patterns of CD4+ naive, Tcm and early TemRO cells, although these effects
263
were not reflected in the total CD4+ memory T-cell populations. In contrast, female gender
264
was associated with a decrease in the patterns of CD4+ intermediate and late TemRA and
265
CD8+ late TemRA cells. Whereas premature rupture of membranes was associated with an
14
ACCEPTED MANUSCRIPT increase in CD4+ late TemRO and CD4+ intermediate TemRA cells, child birth via caesarian
267
section was associated with a reduction in the pattern of CD4+ early TemRO cells. Persistent
268
viral infection with CMV and EBV significantly affected CD4+ and CD8+ T-cell dynamics, in
269
contrast to breastfeeding, antibiotics usage, H. pylori, HSV-1 and VZV carriership. CMV
270
infection associated exclusively with an increase in the patterns of intermediate and late
271
differentiated CD4+ and CD8+ TemRO and TemRA cells. Associations with EBV were
272
slightly more variable, also including early differentiated CD4+ and CD8+ TemRO cells.
M AN US C
RI PT
266
Within the TCRγδ+ T-cell lineage, premature rupture of membranes was associated with
274
an increase in Vδ2+Vγ9+ T cells, CMV infection associated with an increase in the pattern of
275
Vδ1+ T cells.
AC C
EP
TE
D
273
15
ACCEPTED MANUSCRIPT DISCUSSION
277
We here modelled the kinetics of 62 leukocyte subsets, and identified distinct patterns of cell
278
numbers in the first 6 years of life for innate leukocytes, naive B and T cells, and memory B
279
and T cells. Unsupervised clustering revealed that leukocyte dynamics between birth and the
280
age of 6 years could be summarized into 4 major profiles, with either (1) early predominance
281
and fast decline to stable numbers, (2) gradual increased in first year followed by a gradual
282
decline, or a (3 and 4) slow increased in first 1-2 year followed by stabilization.
M AN US C
RI PT
276
Consistent with their early predominance, innate leukocyte kinetics was affected by
284
maternal education level, which might already influence the fetus prenatally. Low maternal
285
education level is strongly related to a less healthy life.34 These observations might suggests
286
that maternal life style is especially important for shaping of innate leukocyte populations,
287
although the exact mediator remains to be determined.
D
283
The non-classical and intermediate monocytes did not consistently cluster with the innate
289
leukocyte subsets due to a later peak in numbers or a slower decline. This altered kinetics
290
could be the result of these being derived from classical monocytes, and not directly from
291
precursors in bone marrow.35 Still, the various subsets of monocytes express distinct levels of
292
proteins involved in HLA-class II-dependent antigen presentation and CD40-CD40L co-
293
stimulation,35 and differ in parasite pattern recognition.35-38 Thus, our data support the need to
294
discriminate intermediate and non-classical monocytes from the dominant population of
295
classical monocytes in kinetics studies.
AC C
EP
TE
288
296
All naive B-cell and T-cell populations followed profile 2 with a gradual increase in the
297
first 14 months followed by a more gradual decline. After the first 1-2 years of life, B-
298
lymphocyte production in bone marrow decreases,39, 40 and the thymus starts to involute.41
299
These processes are likely causes of the decline we observed.
16
ACCEPTED MANUSCRIPT The memory B-cell populations showed different patterns. Natural effector and CD27-
301
IgA+ memory B-cell kinetics were similar to naive B cells. Both memory B-cell populations
302
are suggested to derive, at least in part, from germinal center-independent responses without
303
T-cell help in the splenic marginal zone and the intestinal lamina propria.33,
304
relatively fast rise in these cell numbers within the first 6 months of life contrasts the kinetics
305
of the other, T-cell dependent, memory B-cell subsets. IgMonly, CD27-IgG+ and CD27+IgA+
306
memory B cells followed profile 3, representing gradual memory formation. The CD27+IgG+
307
population followed profile 4, and developed slightly slower than the other memory B-cell
308
populations. These observations supports the concept of rapid generation of T-cell
309
independent memory B cells in the absence of the extensive proliferation and selection
310
processes of the germinal center,33 whereas CD27+IgG+ memory B cells might represent the
311
most mature memory B-cell population. In adults, this subset shows a higher level of affinity
312
maturation, more extensive replication history, and more frequent development via
313
consecutive class-switching, than CD27-IgG+ memory B cells.33, 44 It is perceivable that many
314
of these CD27+IgG+ cells are generated from CD27-IgG+ memory B cells in secondary
315
responses, which could explain the gradual decline in CD27-IgG+ B cell numbers after 14
316
months of age. Together, the different clustering of the memory B-cell populations thereby
317
follows their increasing functional maturity (Suppl Figure 5I).
RI PT
300
The
AC C
EP
TE
D
M AN US C
42, 43
318
Several external determinants significantly affected the B-cell kinetics. First, we
319
confirmed our previous observations regarding lower numbers of memory B cells in breastfed
320
children.20, 22 However, in the current analysis we could not reproduce the previously reported
321
association between breastfeeding duration and germinal center-dependent memory B-cell
322
numbers at 6 months of age.22 The difference is likely due to our current analysis in which
323
duration of breastfeeding was not included and the overall pattern between birth and 6 years
324
of age was studied.
17
ACCEPTED MANUSCRIPT IgA+ memory B cells were significantly increased in H. pylori positive children. The
326
colonization of the gastric mucus by H. pylori was found to correlate with an increase in total
327
blood B-cell counts,45 on top of the local expansion of antibody-secreting IgA+ cells,46 which
328
are important for the protection against H. pylori infection.47 Our observed expansion of
329
circulating IgA+ memory B cells included both the CD27+IgA+ and mucosa-derived CD27-
330
IgA+ subsets. This would suggest that the presence of H. pylori does not only result in a local
331
expansion of plasma cells, but also a systemic expansion of memory B cells in otherwise
332
asymptomatic carriers. It remains to be determined whether this expansion is beneficial for
333
the host as the bacterial protein CagA inhibits B-cell apoptosis and thereby increases the risk
334
for mucosa-associated B-cell malignancies.48, 49
M AN US C
RI PT
325
Memory T-cell subsets were found in clusters 2, 3 and 4. CD4+ and CD8+ Tcm (cluster
336
2) express lymph node homing markers and are the presumed precursors of effector memory
337
T cells.24, 50, 51 The early peak in Tcm numbers prior to those of effector memory T cells in
338
young children would fit with this function. Similarly, CD8+ early TemRA cells might be
339
precursors for further differentiated TemRA subsets. Vδ1+ T-cell numbers (cluster 2) peaked
340
prior to the Vδ2+ and Vγ9+ subsets (cluster 4), thereby confirming the previously observed
341
early shift from Vδ1+ to Vδ2+Vγ9+ predominance in children,52, 53
AC C
EP
TE
D
335
342
In line with previous observations in both children and elderly,22, 54-58 both CD4+ and
343
CD8+ memory T-cell dynamics were predominantly affected by CMV and/or EBV
344
seropositivity. This concerned mostly late effector memory T-cell numbers for CMV and
345
early effector memory T-cell subsets for EBV.24, 59 HSV-1 seropositivity did not affect CD4+
346
and CD8+ T-cell populations,22,
347
association has, to our knowledge, not been described before. EBV infection was associated
348
with a reduction in memory B cells, likely due to the selective EBV persistence in these
349
cells.60, 61 Our large-scale analysis allowed us to separate the herpesvirus-associated effects,
58
but was associated with a loss of naive B cells. This
18
ACCEPTED MANUSCRIPT 350
with CMV and EBV being associated with memory T-cell expansions, and HSV-1 and EBV,
351
with a decrease in naive or memory B-cell numbers, respectively. Gender had a widespread effect on 14 subsets within multiple leukocyte lineages.
353
Interestingly, girls showed a skewing of humoral and early differentiated CD4+ T-cell
354
responses over cellular cytotoxic responses, in contrast to boys. These effects might be
355
associated with differences in sex-hormone levels (testosterone, estradiol) that are already
356
detectable during early infancy, as well as with genetic differences between females and
357
males.62 These insight can especially be important for dissection of auto-immune diseases,
358
which are much more prevalent in females than in males, though predisposition for allergic
359
diseases seems to be opposite in infancy.62, 63
M AN US C
RI PT
352
360
Methodological considerations
362
The strength of this study is its prospective longitudinal population-based design with >1,000
363
children and the possibility to study 26 external determinants in 4 subgroups with
364
adjustments for major confounders. The linear mixed model approach enabled modeling of
365
cross-sectional data, with further improvement of the accuracy by incorporating additionally
366
available longitudinal follow-up data. Furthermore, we included only children with a two-
367
generation Dutch ancestry, which prevented interference of our analyses by strong ethnic and
368
cultural influences. However, extrapolation of our findings to different ethnic and cultural
369
populations might be limited, and would require additional analysis of ethnically-different
370
population cohorts.
AC C
EP
TE
D
361
371
The inclusion of 62 leukocyte populations allowed for the large-scale analysis of the
372
effect of external determinants on both total cell lineages as well as on small subsets defined
373
by extensive and detailed phenotypic definition. The overlap in some populations could have
374
skewed the hierarchical clustering. However, our selection of 31 phenotypically non-
19
ACCEPTED MANUSCRIPT overlapping populations resulted in clusters with similar patterns, indicating the robustness of
376
the 4 major patterns of leukocyte dynamics and the observed effects of external determinants.
377
Our study was primarily explorative with a focus on the identification of determinants
378
that affected leukocyte dynamics between birth and 6 years of age. Although we defined
379
whether determinants had a positive or negative association with leukocyte numbers, we were
380
unable to identify the exact nature of the effect, i.e. exactly when these effects presented and
381
whether these effects will be transient, persistent or potentially even increasing over time.
382
Consequently, more research into individual determinants will be needed to extend our
383
observations by specifying these effects.
M AN US C
RI PT
375
Finally, special consideration needs to be taken for serology of infectious agents that
385
were measured at the age of 6 years. We cannot determine the exact timing of primary
386
infections, and this can consequently be in the whole period preceding the age of 6 years.
387
Still, most H. pylori infections already occur in early childhood, and IgG seropositivity to
388
herpesviruses only appears several weeks to 3 months after infection. Thus, these
389
determinants were present already during or prior to the 6th year of life.
390
EP
TE
D
384
Conclusions
392
With our unbiased approach, we here, for the first time, classified leukocyte populations
393
according to their dynamics between birth and 6 years of age. Moreover, we identified
394
nongenetic factors that are associated with the dynamics of cell lineages or specific leukocyte
395
subsets and underlie, at least in part, the human immunological diversity. These newly
396
identified determinants can provide new targets for studies on the molecular processes that
397
regulate leukocyte development and immune responses, and that together underlie formation
398
of long-lasting immunity without inducing destructive, excessive or insufficient immune
399
responses.
AC C
391
20
ACCEPTED MANUSCRIPT ACKNOWLEDGEMENTS
401
This work was supported by an Erasmus MC Fellowship to M.C.v.Z, and was performed
402
within the framework of the Erasmus MC Postgraduate School Molecular Medicine. The
403
Generation R Study is conducted by the Erasmus MC, Erasmus University Rotterdam in
404
close collaboration with the School of Law and Faculty of Social Sciences of the Erasmus
405
University Rotterdam, the Municipal Health Service Rotterdam Metropolitan Area, the
406
Rotterdam Homecare Foundation, and the Stichting Trombosedienst & Artsenlaboratorium
407
Rijnmond. We gratefully acknowledge the contributions of children and parents, general
408
practitioners, hospitals, midwives and pharmacies in Rotterdam. We thank D. Zhao, K.A.M.
409
van Kester, M.A.W. Smits-te Nijenhuis, M.J Kolijn-Couwenberg and N.M.A. Nagtzaam for
410
technical support.
M AN US C
RI PT
400
D
411
CONFLICT OF INTEREST
413
All authors declare that no competing interests exist.
414
J.J.M.v.D., E.G.v.L., L.E.B-J, H.H., W.A.D., and M.C.v.Z. designed and supervised the flow
415
cytometry experiments; D.v.d.H., M.A.E.J. and H.B.-C. performed and analyzed most of the
416
experiments and contributed to data analyses; V.W.V.J. designed the Generation R Study;
417
H.A.M. designed the study, the data collection and the analyses of determinant data; K.N.
418
contributed to the statistical modeling; D.v.d.H., M.C.v.Z., and M.A.E.J. wrote the
419
manuscript; and all authors commented on the manuscript.
AC C
EP
TE
412
21
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
RI PT
REFERENCES 1. Comans-Bitter WM, de Groot R, van den Beemd R, Neijens HJ, Hop WC, Groeneveld K, et al. Immunophenotyping of blood lymphocytes in childhood. Reference values for lymphocyte subpopulations. J Pediatr 1997; 130:388-93. 2. van Gent R, van Tilburg CM, Nibbelke EE, Otto SA, Gaiser JF, Janssens-Korpela PL, et al. Refined characterization and reference values of the pediatric T- and B-cell compartments. Clin Immunol 2009; 133:95-107. 3. Tosato F, Bucciol G, Pantano G, Putti MC, Sanzari MC, Basso G, et al. Lymphocytes subsets reference values in childhood. Cytometry A 2015; 87:81-5. 4. Duchamp M, Sterlin D, Diabate A, Uring-Lambert B, Guerin-El Khourouj V, Le Mauff B, et al. B-cell subpopulations in children: National reference values. Immun Inflamm Dis 2014; 2:131-40. 5. Xanthou M. Leucocyte blood picture in healthy full-term and premature babies during neonatal period. Arch Dis Child 1970; 45:242-9. 6. Manroe BL, Weinberg AG, Rosenfeld CR, Browne R. The neonatal blood count in health and disease. I. Reference values for neutrophilic cells. J Pediatr 1979; 95:89-98. 7. Sagnia B, Ateba Ndongo F, Ndiang Moyo Tetang S, Ndongo Torimiro J, Cairo C, Domkam I, et al. Reference values of lymphocyte subsets in healthy, HIV-negative children in Cameroon. Clin Vaccine Immunol 2011; 18:790-5. 8. Huck K, Feyen O, Ghosh S, Beltz K, Bellert S, Niehues T. Memory B-cells in healthy and antibody-deficient children. Clin Immunol 2009; 131:50-9. 9. Morbach H, Eichhorn EM, Liese JG, Girschick HJ. Reference values for B cell subpopulations from infancy to adulthood. Clin Exp Immunol 2010; 162:271-9. 10. Piatosa B, Wolska-Kusnierz B, Pac M, Siewiera K, Galkowska E, Bernatowska E. B cell subsets in healthy children: reference values for evaluation of B cell maturation process in peripheral blood. Cytometry B Clin Cytom 2010; 78:372-81. 11. Driessen GJ, Dalm VA, van Hagen PM, Grashoff HA, Hartwig NG, van Rossum AM, et al. Common variable immunodeficiency and idiopathic primary hypogammaglobulinemia: two different conditions within the same disease spectrum. Haematologica 2013; 98:1617-23. 12. Brodin P, Jojic V, Gao T, Bhattacharya S, Angel CJ, Furman D, et al. Variation in the human immune system is largely driven by non-heritable influences. Cell 2015; 160:3747. 13. Orru V, Steri M, Sole G, Sidore C, Virdis F, Dei M, et al. Genetic variants regulating immune cell levels in health and disease. Cell 2013; 155:242-56. 14. Carr EJ, Dooley J, Garcia-Perez JE, Lagou V, Lee JC, Wouters C, et al. The cellular composition of the human immune system is shaped by age and cohabitation. Nat Immunol 2016; 17:461-8. 15. Pabst HF, Spady DW, Pilarski LM, Carson MM, Beeler JA, Krezolek MP. Differential modulation of the immune response by breast- or formula-feeding of infants. Acta Paediatr 1997; 86:1291-7. 16. Winkler B, Aulenbach J, Meyer T, Wiegering A, Eyrich M, Schlegel PG, et al. Formulafeeding is associated with shift towards Th1 cytokines. Eur J Nutr 2015; 54:129-38. 17. Hanson L, Silfverdal SA, Stromback L, Erling V, Zaman S, Olcen P, et al. The immunological role of breast feeding. Pediatr Allergy Immunol 2001; 12 Suppl 14:15-9. 18. Hanson LA, Korotkova M, Lundin S, Haversen L, Silfverdal SA, Mattsby-Baltzer I, et al. The transfer of immunity from mother to child. Ann N Y Acad Sci 2003; 987:199206.
22
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
RI PT
19. Duijts L, Ramadhani MK, Moll HA. Breastfeeding protects against infectious diseases during infancy in industrialized countries. A systematic review. Matern Child Nutr 2009; 5:199-210. 20. Hawkes JS, Neumann MA, Gibson RA. The effect of breast feeding on lymphocyte subpopulations in healthy term infants at 6 months of age. Pediatr Res 1999; 45:648-51. 21. M'Rabet L, Vos AP, Boehm G, Garssen J. Breast-feeding and its role in early development of the immune system in infants: consequences for health later in life. J Nutr 2008; 138:1782S-90S. 22. Jansen MA, van den Heuvel D, van Zelm MC, Jaddoe VW, Hofman A, de Jongste JC, et al. Decreased memory B cells and increased CD8 memory T cells in blood of breastfed children: the generation R study. PLoS One 2015; 10:e0126019. 23. Bekker V, Scherpbier H, Pajkrt D, Jurriaans S, Zaaijer H, Kuijpers TW. Persistent humoral immune defect in highly active antiretroviral therapy-treated children with HIV1 infection: loss of specific antibodies against attenuated vaccine strains and natural viral infection. Pediatrics 2006; 118:e315-22. 24. Appay V, van Lier RA, Sallusto F, Roederer M. Phenotype and function of human T lymphocyte subsets: consensus and issues. Cytometry A 2008; 73:975-83. 25. Jaddoe VW, van Duijn CM, Franco OH, van der Heijden AJ, van Iizendoorn MH, de Jongste JC, et al. The Generation R Study: design and cohort update 2012. Eur J Epidemiol 2012; 27:739-56. 26. Kruithof CJ, Kooijman MN, van Duijn CM, Franco OH, de Jongste JC, Klaver CC, et al. The Generation R Study: Biobank update 2015. Eur J Epidemiol 2014; 29:911-27. 27. Kalina T, Flores-Montero J, van der Velden VH, Martin-Ayuso M, Bottcher S, Ritgen M, et al. EuroFlow standardization of flow cytometer instrument settings and immunophenotyping protocols. Leukemia 2012; 26:1986-2010. 28. Korevaar TI, Medici M, de Rijke YB, Visser W, de Muinck Keizer-Schrama SM, Jaddoe VW, et al. Ethnic differences in maternal thyroid parameters during pregnancy: the Generation R study. J Clin Endocrinol Metab 2013; 98:3678-86. 29. Kiefte-de Jong JC, Jaddoe VW, Uitterlinden AG, Steegers EA, Willemsen SP, Hofman A, et al. Levels of antibodies against tissue transglutaminase during pregnancy are associated with reduced fetal weight and birth weight. Gastroenterology 2013; 144:72635 e2. 30. Jansen MA, Tromp, II, Kiefte-de Jong JC, Jaddoe VW, Hofman A, Escher JC, et al. Infant feeding and anti-tissue transglutaminase antibody concentrations in the Generation R Study. Am J Clin Nutr 2014; 100:1095-101. 31. den Hollander WJ, Holster IL, van Gilst B, van Vuuren AJ, Jaddoe VW, Hofman A, et al. Intergenerational reduction in Helicobacter pylori prevalence is similar between different ethnic groups living in a Western city. Gut 2014. 32. R Core Team; R Foundation for Statistical Computing V, Austria. R: A language and environment for statistical computing. . 2015. 33. Berkowska MA, Driessen GJ, Bikos V, Grosserichter-Wagener C, Stamatopoulos K, Cerutti A, et al. Human memory B cells originate from three distinct germinal centerdependent and -independent maturation pathways. Blood 2011; 118:2150-8. 34. Bouthoorn SH, Silva LM, Murray SE, Steegers EA, Jaddoe VW, Moll H, et al. Loweducated women have an increased risk of gestational diabetes mellitus: the Generation R Study. Acta Diabetol 2015; 52:445-52. 35. Wong KL, Tai JJ, Wong WC, Han H, Sem X, Yeap WH, et al. Gene expression profiling reveals the defining features of the classical, intermediate, and nonclassical human monocyte subsets. Blood 2011; 118:e16-31.
23
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
RI PT
36. Turner JD, Bourke CD, Meurs L, Mbow M, Dieye TN, Mboup S, et al. Circulating CD14brightCD16+ 'intermediate' monocytes exhibit enhanced parasite pattern recognition in human helminth infection. PLoS Negl Trop Dis 2014; 8:e2817. 37. Hijdra D, Vorselaars AD, Grutters JC, Claessen AM, Rijkers GT. Phenotypic characterization of human intermediate monocytes. Front Immunol 2013; 4:339. 38. Ziegler-Heitbrock L, Hofer TP. Toward a refined definition of monocyte subsets. Front Immunol 2013; 4:23. 39. Kogut I, Scholz JL, Cancro MP, Cambier JC. B cell maintenance and function in aging. Semin Immunol 2012; 24:342-9. 40. Jensen K, Schaffer L, Olstad OK, Bechensteen AG, Hellebostad M, Tjonnfjord GE, et al. Striking decrease in the total precursor B-cell compartment during early childhood as evidenced by flow cytometry and gene expression changes. Pediatr Hematol Oncol 2010; 27:31-45. 41. Flores KG, Li J, Sempowski GD, Haynes BF, Hale LP. Analysis of the human thymic perivascular space during aging. J Clin Invest 1999; 104:1031-9. 42. Weller S, Braun MC, Tan BK, Rosenwald A, Cordier C, Conley ME, et al. Human blood IgM "memory" B cells are circulating splenic marginal zone B cells harboring a prediversified immunoglobulin repertoire. Blood 2004; 104:3647-54. 43. Weller S, Faili A, Garcia C, Braun MC, Le Deist FF, de Saint Basile GG, et al. CD40CD40L independent Ig gene hypermutation suggests a second B cell diversification pathway in humans. Proc Natl Acad Sci U S A 2001; 98:1166-70. 44. Jackson KJ, Wang Y, Collins AM. Human immunoglobulin classes and subclasses show variability in VDJ gene mutation levels. Immunol Cell Biol 2014; 92:729-33. 45. Helmin-Basa A, Michalkiewicz J, Gackowska L, Kubiszewska I, Eljaszewicz A, Mierzwa G, et al. Pediatric Helicobacter pylori infection and circulating T-lymphocyte activation and differentiation. Helicobacter 2011; 16:27-35. 46. Mattsson A, Quiding-Jarbrink M, Lonroth H, Hamlet A, Ahlstedt I, Svennerholm A. Antibody-secreting cells in the stomachs of symptomatic and asymptomatic Helicobacter pylori-infected subjects. Infect Immun 1998; 66:2705-12. 47. Lee CK, Weltzin R, Thomas WD, Jr., Kleanthous H, Ermak TH, Soman G, et al. Oral immunization with recombinant Helicobacter pylori urease induces secretory IgA antibodies and protects mice from challenge with Helicobacter felis. J Infect Dis 1995; 172:161-72. 48. Umehara S, Higashi H, Ohnishi N, Asaka M, Hatakeyama M. Effects of Helicobacter pylori CagA protein on the growth and survival of B lymphocytes, the origin of MALT lymphoma. Oncogene 2003; 22:8337-42. 49. Lin WC, Tsai HF, Kuo SH, Wu MS, Lin CW, Hsu PI, et al. Translocation of Helicobacter pylori CagA into Human B lymphocytes, the origin of mucosa-associated lymphoid tissue lymphoma. Cancer Res 2010; 70:5740-8. 50. Sallusto F, Geginat J, Lanzavecchia A. Central memory and effector memory T cell subsets: function, generation, and maintenance. Annu Rev Immunol 2004; 22:745-63. 51. Sallusto F, Lenig D, Forster R, Lipp M, Lanzavecchia A. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 1999; 401:708-12. 52. Sandberg Y, Almeida J, Gonzalez M, Lima M, Barcena P, Szczepanski T, et al. TCRgammadelta+ large granular lymphocyte leukemias reflect the spectrum of normal antigen-selected TCRgammadelta+ T-cells. Leukemia 2006; 20:505-13. 53. Breit TM, Wolvers-Tettero IL, van Dongen JJ. Unique selection determinant in polyclonal V delta 2-J delta 1 junctional regions of human peripheral gamma delta T lymphocytes. J Immunol 1994; 152:2860-4. 24
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
RI PT
54. Bekker V, Bronke C, Scherpbier HJ, Weel JF, Jurriaans S, Wertheim-van Dillen PM, et al. Cytomegalovirus rather than HIV triggers the outgrowth of effector CD8+CD45RA+CD27- T cells in HIV-1-infected children. Aids 2005; 19:1025-34. 55. Weinberger B, Lazuardi L, Weiskirchner I, Keller M, Neuner C, Fischer KH, et al. Healthy aging and latent infection with CMV lead to distinct changes in CD8+ and CD4+ T-cell subsets in the elderly. Hum Immunol 2007; 68:86-90. 56. van Leeuwen EM, Remmerswaal EB, Vossen MT, Rowshani AT, Wertheim-van Dillen PM, van Lier RA, et al. Emergence of a CD4+CD28- granzyme B+, cytomegalovirusspecific T cell subset after recovery of primary cytomegalovirus infection. J Immunol 2004; 173:1834-41. 57. Chidrawar S, Khan N, Wei W, McLarnon A, Smith N, Nayak L, et al. Cytomegalovirusseropositivity has a profound influence on the magnitude of major lymphoid subsets within healthy individuals. Clin Exp Immunol 2009; 155:423-32. 58. Derhovanessian E, Maier AB, Hahnel K, Beck R, de Craen AJ, Slagboom EP, et al. Infection with cytomegalovirus but not herpes simplex virus induces the accumulation of late-differentiated CD4+ and CD8+ T-cells in humans. J Gen Virol 2011; 92:2746-56. 59. Appay V, Dunbar PR, Callan M, Klenerman P, Gillespie GM, Papagno L, et al. Memory CD8+ T cells vary in differentiation phenotype in different persistent virus infections. Nat Med 2002; 8:379-85. 60. Babcock GJ, Decker LL, Volk M, Thorley-Lawson DA. EBV persistence in memory B cells in vivo. Immunity 1998; 9:395-404. 61. Chaganti S, Heath EM, Bergler W, Kuo M, Buettner M, Niedobitek G, et al. EpsteinBarr virus colonization of tonsillar and peripheral blood B-cell subsets in primary infection and persistence. Blood 2009; 113:6372-81. 62. Ober C, Loisel DA, Gilad Y. Sex-specific genetic architecture of human disease. Nat Rev Genet 2008; 9:911-22. 63. Whitacre CC, Reingold SC, O'Looney PA. A gender gap in autoimmunity. Science 1999; 283:1277-8.
25
ACCEPTED MANUSCRIPT Table 1. Definition of leukocyte subsets
+ + + + + + + + + + + + + +
SSClowCD45+CD3+ SSClowCD3+TCRγδ-TCRαβ+ SSClowCD3+TCRγδ-TCRαβ+CD8-CD4+ SSClowCD3+TCRγδ-TCRαβ+CD8+CD4SSClowCD3+TCRγδ+TCRαβSSClowCD3+TCRγδ+TCRαβ-CD8-CD4+ SSClowCD3+TCRγδ+TCRαβ-CD8+CD4SSClowCD3+TCRαβ-Vδ1-Vδ2SSClowCD3+TCRαβ-Vδ1-Vδ2+ SSClowCD3+TCRαβ-Vδ1+Vδ2FSClowSSClowCD3+TCRαβ-Vγ9+ FSClowSSClowCD3+TCRαβ-Vδ2+Vγ9+ SSClowCD45+CD3+CD4+CD8FSClowSSClowCD3+CD8-CD197+CD45RO-CD27+CD28+ FSClowSSClowCD3+CD8- CD197- & CD197+CD45RO+ FSClowSSClowCD3+CD8-CD197+CD45RO+CD27+CD28+ FSClowSSClowCD3+CD8-CD197-CD45RO+ FSClowSSClowCD3+CD8-CD197-CD45RO+CD27+CD28+ FSClowSSClowCD3+CD8-CD197-CD45RO+CD27-CD28+ FSClowSSClowCD3+CD8-CD197-CD45RO+CD27-CD28FSClowSSClowCD3+CD8-CD197-CD45ROFSClowSSClowCD3+CD8-CD197-CD45RO-CD27+CD28+ FSClowSSClowCD3+CD8-CD197-CD45RO-CD27-CD28+ FSClowSSClowCD3+CD8-CD197-CD45RO-CD27-CD28SSClowCD45+CD3+CD4-CD8+ FSClowSSClowCD3+CD8+CD197+CD45RO-CD27+CD28+ FSClowSSClowCD3+CD8+ CD197- & CD197+CD45RO+ FSClowSSClowCD3+CD8+CD197+CD45RO+CD27+CD28+ FSClowSSClowCD3+CD8+CD197-CD45RO+ FSClowSSClowCD3+CD8+CD197-CD45RO+CD27+CD28+ FSClowSSClowCD3+CD8+CD197-CD45RO+CD27+CD28FSClowSSClowCD3+CD8+CD197-CD45RO+CD27-CD28FSClowSSClowCD3+CD8+CD197-CD45ROFSClowSSClowCD3+CD8+CD197-CD45RO-CD27+CD28+ FSClowSSClowCD3+CD8+CD197-CD45RO-CD27+CD28FSClowSSClowCD3+CD8+CD197-CD45RO-CD27-CD28SSClowCD45+CD19+ SSClowCD19+CD27-IgD+ SSClowCD19+ IgD- & CD27+IgD+ SSClowCD19+CD27-IgDSSClowCD19+CD27+IgDSSClowCD19+CD27+IgM+ SSClowCD19+CD27+IgM+IgD+ SSClowCD19+CD27+IgM+IgDSSClowCD19+IgA+ SSClowCD19+CD27-IgA+ SSClowCD19+CD27+IgA+ SSClowCD19+IgG+ SSClowCD19+CD27-IgG+ SSClowCD19+CD27+IgG+ SSClowCD19+CD38lowCD21low SSClowCD19+Igκ-Igλ+ SSClowCD19+Igκ+Igλ-
+ + + +
+ + + + + +
RI PT
Lymphocytes T cells TCRαβ+ T cells ∟CD4+ TCRαβ+ T cells ∟CD8+ TCRαβ+ T cells TCRγδ+ T cells ∟CD4+ TCRγδ+ T cells ∟CD8+ TCRγδ+ T cells ∟Vδ1-Vδ2- TCRαβ- T cells ∟Vδ2+ T cells ∟Vδ1+ T cells ∟Vγ9+ T cells ∟Vδ2+Vγ9+ T cells CD4+ T cells ∟CD4+ Tnaive ∟CD4+ Tmem ∟CD4+ Tcm ∟CD4+ TemRO ∟CD4+ early TemRO ∟ CD4+ interm TemRO ∟CD4+ late TemRO ∟CD4+ TemRA ∟CD4+ early TemRA ∟ CD4+ interm TemRA ∟CD4+ late TemRA CD8+ T cells ∟CD8+ Tnaive ∟CD8+ Tmem ∟CD8+ Tcm ∟CD8+ TemRO ∟CD8+ early TemRO ∟CD8+ interm TemRO ∟CD8+ late TemRO ∟CD8+ TemRA ∟CD8+ early TemRA ∟CD8+ interm TemRA ∟CD8+ late TemRA B cells Bnaive Bmem ∟CD27- Bmem ∟CD27+ Bmem ∟IgM+ Bmem ∟Natural effector ∟IgMonly ∟IgA+ Bmem ∟CD27-IgA+ ∟CD27+IgA+ ∟IgG+ Bmem ∟CD27-IgG+ ∟CD27+IgG+ CD21low B cells Igλ+ B cells Igκ+ B cells
M AN US C
+
SSChighCD45+ SSChighCD45+CD15+ SSChighCD45+CD16+ SSChighCD45+CD16FSCinterSSCinterCD45+ FSCinterSSCinterCD45+CD14+CD16FSCinterSSCinterCD45+CD14+CD16+ FSCinterSSCinterCD45+CD14-CD16+ SSClowCD45+CD3-CD16+or CD56+
D
+ + + +
Phenotype definition
TE
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62
+ +
Name population Innate leukocytes Granulocytes CD15+ granulocytes Neutrophils Eosinophils Monocytes Classical monocytes Intermediate monocytes Non-classical monocytes NK cells
EP
1 2 3 4 5 6 7 8 9
n=31
AC C
n=62
26
Table 2. Characteristics of nongenetic factors negative individuals (%)
IP T
ACCEPTED MANUSCRIPT
positive individuals (%)
missing (%)
significant effect on cluster a
AC C
EP
TE
D
M AN US
CR
Prenatal maternal life style Maternal age (years) (continuous determinant) 0 (0) Low maternal educational level 400 (33.8) 760 (64.3) 22 (1.9) 1,2,3,4 Net household income per month >€ 2,200 242 (20.5) 847 (71.7) 93 (7.9) Smoking during pregnancy 815 (69) 256 (21.7) 111 (9.4) Alcohol use continued during pregnancy 299 (25.3) 750 (63.5) 133 (11.3) Body Mass Index before pregnancy (kg/m2) (continuous determinant) 181 (15.3) Prenatal maternal immune-mediated diseases Maternal atopy (eczema, allergy HDM, hay-fever 683 (57.8) 383 (32.4) 116 (9.8) Anti-TPO (before 18 weeks of pregnancy) (mU/ml) (continuous determinant) 278 (23.5) 2 Maternal TG2A during pregnancy U/ml (continuous determinant) 208 (17.6) Birth characteristics Gender (girl yes/no) 600 (50.8) 582 (49.2) 0 (0) 1,2,3,4 Preterm birth <37 weeks 76 (6.4) 1,106 (93.6) 0 (0) Low birth weight <2,500g 1,119 (94.7) 63 (5.3) 0 (0) Premature rupture of membranes 1,089 (92.1) 53 (4.5) 40 (3.4) 4 Caesarian section versus vaginal/ forceps/vacuum assisted 903 (76.4) 154 (13) 125 (10.6) 3,4 Birth season (born in Fall/Winter) 682 (57.7) 500 (42.3) 0 (0) Bacterial/viral exposure-related characteristics Breastfeeding ever 109 (9.2) 969 (82) 104 (8.8) Breastfeeding at 6 months of age 754 (63.8) 299 (25.3) 129 (10.9) 2,4 Siblings >1 687 (58.1) 480 (40.6) 15 (1.3) 1 Antibiotics/Penicillin use in 1st y 636 (53.8) 342 (28.9) 204 (17.3) 1,3 Upper respiratory tract infections in 1st y 547 (46.3) 475 (40.2) 160 (13.5) Lower respiratory tract infections in 1st y (doctor attended) 819 (69.3) 131 (11.1) 232 (19.6) Carrier of Helicobacter pylori within 6yrs 887 (75) 49 (4.1) 246 (20.8) 2,3 Seropositivity for Cytomegalovirus (CMV) at 6y 665 (56.3) 269 (22.8) 248 (21) 1,2,3,4 Seropositivity for Epstein Barr virus (EBV) at 6y 534 (45.2) 400 (33.8) 248 (21) 2,3,4 Seropositivity for Herpes simplex virus-1 (HSV-1) at 6y 810 (68.5) 124 (10.5) 248 (21) 2 Seropositivity for Varicella zoster virus (VZV) at 6y 73 (6.2) 861 (72.8) 248 (21) a First, each determinant was added to the linear mixed effect model univariably to test for a significant effect (Significance was defined as p<0.0125); Subsequently, per leukocyte population, all determinants with a significant individual effect were added in a multivariable model to correct for confounding effects. Up to 3 determinants were added per multivariable model. Significance was defined as p<0.05. IU, international units. No confounding effect was observed
27
ACCEPTED MANUSCRIPT Figure Legends (n=4)
in children between birth and 6 years of age.
RI PT
Figure 1. Dynamics of innate leukocyte and naive and memory lymphocyte populations
Dynamics of A) monocytes, neutrophils, eosinophils and NK cells, B) naive and memory B-
M AN US C
cell and T-cell subsets, and C) TCRγδ+ T cells. Flow cytometry plots depict population definitions in one representative 6-year-old individual. Graphs depict absolute numbers of cells in blood of 1,182 children with in total 2,010 measurements (gray dots). Linear mixed effect models are depicted by a solid black line for each population and the 90% confidence interval (CI) of the model with dashed black lines. For clarity, only direct follow-up time
D
points of an individual were connected with gray lines; i.e. 0-6m, 6-14m, 14-25m or 25-76m.
TE
Figure 2. Hierarchical clustering of leukocyte subset dynamics in early childhood. A) Ward’s hierarchical clustering of the cellularity of the 62 populations derived from the
EP
linear mixed models. Cellularity between 0-76 months of age was converted to zero mean
AC C
and unit standard deviation (z score). B) Average patterns +/- 1 standard deviation of the subsets in each of the 4 major clusters. Populations are numbered according to their order in Table 1.
Figure 3. Nongenetic factors associated with leukocyte dynamics in the four clusters. Significant associations of the 26 determinants with leukocyte subsets in clusters 1-4 based on the clustering of 62 leukocyte populations as in Figure 2 (A and C) or of the 31 nonoverlapping subsets as in Suppl Figure 6 (B and D) corrected for confounding effects. A and B) Relative distribution of types of determinants that significantly affect leukocyte kinetics. C and D) Frequencies of populations within a cluster that were affected by individual 28
ACCEPTED MANUSCRIPT determinants. Abbreviations: PROMM, premature rupture of membranes; RTI, respiratory
RI PT
tract infection. Numbers in each bar refer to the populations as defined in Table 1.
Figure 4. Effects of nongenetic factors on the dynamics of innate leukocytes, and B-cell and T-cell subsets.
M AN US C
Significant effects are shown derived from multivariable linear mixed effect analysis of all 26 determinants on the 31 non-overlapping populations, and the total memory subsets (“mem”) within the B-cell, CD4+ T-cell and CD8+ T-cell lineages. Blue boxes represent a reduction, red boxes an increase in the pattern of the indicated population. Effect sizes are presented in Suppl Table 3, and details of each determinant in Table 2. Additional abbreviations: PROMM, premature rupture of membranes; RTI, respiratory tract infection; neu, neutrophils;
D
eos, eosinophils; class, classical monocytes; inter, intermediate monocytes; n-class, non-
AC C
EP
TE
classical monocytes.
29
B lympho
monocytes
3000 2000
Tnaive
2000 1000 500 0
0
D
age (months)
Tnaive
cells/µl
CD45RO
TE
12 24 36 48 60 72 84 96
AC C
CD3
T cells
TCRγδ+ T cells
1500
1000 500
12 24 36 48 60 72 84 96
CD4+ Tnaive
0
2500
12 24 36 48 60 72 84 96
CD4+ Tmem
2000 1500 1000 500 0
0
12 24 36 48 60 72 84 96
age (months)
1000 800 600 400 200
0
CD8+ Tmem
2000
0
0
9000 8000 7000 6000 5000 4000 3000 2000 1000 0
0
TCRαβ
12 24 36 48 60 72 84 96
1500
1000
TCRγδ+ T cells
1200
0
1500
1400
cells/μl
2500
C
EP
NK cells
3000
Tmem
CCR7
12 24 36 48 60 72 84 96
3500
cells/µl
NK cells
CD4+
CD3
0
lymphocytes
CCR7 CD4+ T cells
CD8+
TCRγδ
CD45
eosinophils
CD8
eosino
9000 8000 7000 6000 5000 4000 3000 2000 1000 0
CD8+ Tnaive
2000
0
12 24 36 48 60 72 84 96
Bmem
2500
500
5000
cells/µl
CD16
T-cells
10000
Tmem
cells/µl
cells/µl
15000
0
12 24 36 48 60 72 84 96
3000 2500
neutrophils
20000
0
2000
0
CD27 CD8+ T cells
CD19
12 24 36 48 60 72 84 96
30000
neutro
3000
0 0
0 25000
granulocytes
800 700 600 500 400 300 200 100 0
Bnaive
1000
1000 0
CD45
CD16/CD56
B cells
1500 500
lympho
5000 4000
CD45RO
mono
2500
CD3
SSC
cells/µl
granulo
B cells Bnaive Bmem
T cells
cells/µl
3500
IgD
all events
M AN U
A
SC
RI PT
ACCEPTED MANUSCRIPT
12 24 36 48 60 72 84 96
age (months)
0
12 24 36 48 60 72 84 96
age (months)
Data of individual children model 90% CI
3
4
0
M AN U
2
SC
4 1 2 3 9 6 5 8 25 52 51 55 62 46 34 47 61 60 13 43 19 54 7 37 17 15 23 10 11 35 22 12 40 38 39 36 41 50 45 42 16 53 24 56 48 58 49 57 31 30 18 28 44 14 32 21 33 20 27 26 29 59
1
10
20
30
D
A
RI PT
ACCEPTED MANUSCRIPT
40
50
Age (months)
TE -2
Z- score
1
0
3.00 2.50 2.00 1.50 1.00 0.50 0.00 -0.50 -1.00 -1.50 -2.00 3.00 2.00 1.00 0.00 -1.00 -2.00 -3.00 -4.00 -5.00
6
2
0
0
1
2
Eosinophils Granulocytes CD15+ granuloytes Neutrophils NK cells Classical monocytes Monocytes Non-classical monocytes
6
12 18 24 30 36 42 48 54 60 66 72
12 18 24 30 36 42 48 54 60 66 72
3
0
2.00
6
4
0.00 -1.00 -2.00 -3.00 -4.00 -5.00
0
6
25 52 51 55 62 46 34 47 61 60 13 43
CD4+ Tcm Natural effector B cells IgM+ Bmem CD27-IgA+ Igκ+ B cells B cells CD8+ T cells Bnaive Igλ+ B cells CD21low B cells CD8+ TCRαβ+ T cells CD8+ TemRA early
19 54 7 37 17 15 23 10 11 35 22 12
Vδ1+ T cells IgA+ Bmem Intermediate monocytes CD8+ Tcm Vδ1-Vδ2- TCRαβ- T cells CD4+ TCRγδ+ T cells CD4+ Tnaive T cells TCRαβ+ T cells CD8+ Tnaive CD4+ T cells CD4+ TCRαβ+ T cells
40 38 39 36 41 50 45 42 16
CD8+ TemRO interm CD8+ TemRO CD8+ TemRO early CD8+ Tmem CD8+ TemRO late CD27+ Bmem CD8+ TemRA late CD8+ TemRA CD8+ TCRγδ+ T cells
53 24 56 48 58 49 57 31 30
IgMonly CD4+ Tmem CD27+IgA+ Bmem CD27-IgG+ CD27- Bmem IgG+ Bmem CD4+ TemRA early CD4+ TemRA
18 28 44 14 32 21 33 20 27
Vδ2+ T cells CD4+ TemRO interm CD8+ TemRA interm TCRγδ+ T cells CD4+ TemRA interm Vδ2+Vγ9+ T cells CD4+ TemRA late Vγ9+ T cells CD4+ TemRO early
26 CD4+ TemRO 29 CD4+ TemRO late 59 CD27+IgG+
12 18 24 30 36 42 48 54 60 66 72
1.00
Z- score
70
EP
4 1 2 3 9 6 5 8
-1
AC C
7.00 6.00 5.00 4.00 3.00 2.00 1.00 0.00 -1.00 -2.00
Z- score
Z- score
B
60
z-score
12 18 24 30 36 42 48 54 60 66 72 Age (months)
A
B
Total 62 populations
31 non-overlapping populations cluster 1
(n=8 effects)
(n=5 effects)
TE
cluster 1 cluster 2
cluster 2
(n= 21 effects)
(n=8 effects)
cluster 3
cluster 3
(n=6 effects)
(n= 24 effects)
EP
cluster 4
cluster 4
(n=20 effects)
10
20
30
40
50
60
70
80
90 100
0
distribution of determinants with a significant effect (%)
Total 62 populations Cluster 1 8 leukocyte pop. 4 5 6
2 3
9
Cluster 2 24 leukocyte pop.
60
37
12 22 23 25
8
9
11 54 55 62
54 55
13 19 34
13 34
16 36 38 40 41 42 56 45
36 38 39 40 41 48 49 50
Cluster 3 18 leukocyte pop.
16 48 49
42 45 56
58
16
46 47 61 62
Cluster 4 12 leukocyte pop. 14 27 32 33 59
18
18 20 21 29 32
26 28 29 33 44 27
26 27 29 33 44
18
age education household income smoking alcohol BMI atopy TPO tTG gender preterm birth low birth weight PROMM sectio birth season breastfeeding ever breastfeeding >6m siblings antibiotics 1y upper RTI 1y lower RTI 1y H.pylori CMV EBV HSV-1 VZV
frequency of populations affected within each cluster (%)
50 40 30 20 10 0 50 40 30 20 10 0 50 40 30 20 10 0 50 40 30 20 10 0
prenatal maternal life style
immunemediated diseases
birth charact.
20
birth characteristics
AC C
maternal immune-mediated
life style
10
30
40
50
60
70
80
90 100
distribution of determinants with a significant effect (%)
bacterial/viral exposure-related characteristics
bacterial/viral exposure
D
31 non-overlapping populations 60 50 40 30 4 20 6 10 0 60 50 40 30 20 10 0 60 50 40 30 20 10 0 60 50 40 30 20 10 0
Cluster 1 4 leukocyte pop.
3
9
9
Cluster 2 11 leukocyte pop.
23 37 25
55
8 55 19
47
Cluster 3 6 leukocyte pop.
56
58 56 40
39 40
Cluster 4 10 leukocyte pop. 27 32 33 21 45 29 59 32 27
28 29 33 41 44 45
27 29 33 41 44.
education TPO gender PROMM sectio breastfeeding >6m siblings antibiotics 1y H.pylori CMV EBV HSV-1
0
frequency of populations affected within each cluster (%)
(n= 23 effects)
C
D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
granulo neu
SC M AN U D
bacterial/viral exposure
TE
birth charact.
age education household income smoking alcohol BMI atopy TPO tTG gender preterm birth low birth weight PROMM sectio birth season breastfeeding ever breastfeeding >6m siblings antibiotics 1y upper RTI 1y lower RTI 1y H. pylori CMV EBV HSV-1 VZV
EP
immunemediated diseases
CD8+ T cells TCRγδ B cells CD4+ T cells IgA+ IgG+ TemRA TemRO TemRO TemRA Nat. IgM Vδ2 NK naive mem eff. only CD27-CD27+ CD27-CD27+ naive mem Tcm early interm late early interm late naive mem Tcm early interm late early interm late Vδ1 Vγ9
AC C
life style
Innate monocytes
eos class inter n-class
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
IP T
Supplemental Table 1. Antibody details
Fluorochrome FITC PE PerCP-Cy5.5 PE-Cy7 APC APC-Cy7 Antibody CD3 CD16+CD56 CD45 CD4 CD19 CD8 Clone SK7 B73.1 + C5.9 2D1 SK3 SJ25C1 SK1 1 Manufacturer BD BD+Dako BD BD BD BD Antibody CD15 CD45 CD16 CD14 Clone MMA 2D1 3G8 MO-P9 2 Manufacturer BD BD BD BD Antibody Igκ Igλ CD19 CD21 Clone polyclonal polyclonal SJ25C1 B-ly-4 3 Manufacturer Dako Southern Biotech BD BD Antibody CD38 CD19 CD21 Clone HB7 SJ25C1 B-ly-4 4 Manufacturer BD BD BD Antibody IgD CD19 IgM CD27 Clone polyclonal SJ25C1 polyclonal L128 5 Manufacturer Southern Biotech BD BD BD Antibody IgA IgG CD19 IgM CD27 Clone polyclonal polyclonal SJ25C1 polyclonal L128 6 Manufacturer Kallestad Southern Biotech BD BD BD Antibody TCRαβ TCRγδ CD3 CD4 CD8 Clone WT31 11F2 SK7 SK3 SK1 7 Manufacturer BD BD BD BD BD Antibody Vδ2 Vδ1 * CD3 CD4 TCRαβ CD8 Clone B6.1 R9.12 SK7 SK3 IP26 SK1 8 Manufacturer BD Beckman Coulter BD BD eBiosciences BD Antibody CD28 CD197 CD3 CD8 CD45RO CD27 Clone CD28.2 3D13 SK7 SK1 UCHL-1 L128 9 Manufacturer BD eBiosciences BD BD BD BD Antibody Vδ2 Vγ9 CD3 CD4 TCRαβ CD8 Clone B6.1 B3.1 SK7 SK3 IP26 SK1 10 Manufacturer BD BD BD BD eBiosciences BD *, unconjugated antibody, detected with Goat anti-Mouse IgG PE (polyclonal; Invitrogen, Waltham, MA) BD, BD Biosciences, San Jose, CA; Dako, DakoCytomation, Glostrup, Denmark; Southern Biotech, Southern Biotech, Birmingham, AL; Kallestad, Kallestad Diagnostics, Chaska, MN; Beckman Coulter, Beckman Coulter, Indianapolis, IN; eBiosciences, eBiosciences, San Diego, CA.
AC C
EP
TE
D
M AN US
CR
Tube
ACCEPTED MANUSCRIPT Supplemental Table 2. Age-associated reference values of leukocyte populations
Age (months)
0 220 0.0 0.0 0.0 0.0 0.0
5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th
4884.2 8263.0 10572.7 13727.9 18784.8 5161.2 8010.4 10397.8 13564.7 17933.3 4433.3 7064.5 9091.8 11634.0 16483.1 489.5 891.3 1340.0 1999.2 3721.9 496.1 1002.2 1346.5 1749.6 2429.0 378.1 829.5 1114.3 1425.4 2125.5 20.6 46.6 72.3 112.1 170.0 24.0 50.6 77.3 116.0 207.5 229.5 460.0 840.0 1350.0 2100.5
Age category (months) 6 14 25 376 241 257 5.5 13.6 23.6 6.0 14.1 24.4 6.2 14.4 25.1 6.7 14.8 25.8 7.7 16.1 27.3
Innate Leukocytes
Neutrophils
EP
AC C
Monocytes
TE
Eosinophils
Classical monocytes
Intermediate monocytes
Non-classical monocytes
NK cells
1196.3 2250.4 3074.1 4143.0 6056.8 1138.2 2072.7 2973.2 4066.6 5633.9 1026.1 1926.4 2750.9 3795.5 5506.1 84.5 178.1 248.0 373.2 763.6 320.6 504.0 691.6 962.7 1510.6 192.7 301.7 425.3 572.6 849.1 26.1 42.4 70.3 116.7 243.3 13.4 24.3 41.3 66.1 139.2 130.0 210.0 280.0 400.0 690.0
1352.5 1990.9 2822.9 3832.6 6467.5 1310.8 1931.9 2690.5 3746.3 6194.1 1116.3 1840.5 2533.5 3665.9 5983.6 83.9 142.2 203.8 315.6 617.5 272.8 433.1 577.4 729.6 1066.6 190.0 304.0 402.6 522.9 714.1 11.6 27.9 43.7 72.0 124.3 4.2 12.6 26.8 48.9 93.4 120.0 180.0 240.0 320.0 536.0
M AN US C
CD15+ granulocytes
D
Granulocytes
1026.1 2055.0 2850.0 4167.8 6519.8 871.6 1735.8 2617.3 3789.1 5913.7 824.4 1668.6 2497.6 3671.1 5653.1 155.2 253.1 351.6 486.7 856.0 360.1 582.8 789.6 1065.8 1755.0 220.4 338.6 461.1 623.9 992.3 19.4 36.0 55.1 87.0 202.7 16.2 39.3 60.8 92.6 162.2 160.0 250.0 330.0 450.0 752.5
72 916 69.3 70.6 72.4 75.0 79.0
RI PT
Number of children
percentile 5th 25th 50th 75th 95th
1562.4 2840.2 3715.0 5041.0 7597.8 1521.0 2736.1 3608.8 4896.7 7527.4 1321.7 2414.2 3274.0 4451.2 6810.9 109.4 199.1 356.4 598.1 1261.5 233.9 370.4 490.9 623.7 924.1 168.8 267.4 349.6 455.9 676.4 10.1 21.0 31.0 44.8 89.8 4.8 11.7 22.0 41.2 80.3 90.0 140.0 190.0 250.3 420.0
ACCEPTED MANUSCRIPT
CD4+TCRαβ+ T cells
CD8+TCRαβ+ T cells
TCRγδ+ T cells
AC C
CD8+ TCRγδ+ T cells
Vδ1-Vδ2- TCRαβ- T cells
Vδ2+ T cells
Vδ1+ T cells
Vγ9+ T cells
2587.5 3740.0 4620.0 5690.0 7722.5 2509.1 3680.3 4477.9 5426.5 7552.0 1788.3 2679.3 3295.1 3994.7 5597.1 575.1 856.5 1101.7 1427.4 2112.0 69.1 123.3 164.0 226.5 333.2 6.6 13.7 20.4 31.6 50.8 8.5 18.2 27.7 44.4 81.0 69.1 113.5 173.8 274.8 448.2 13.1 30.4 43.5 65.5 124.4 29.0 48.2 65.1 96.4 163.8 25.0 39.9 56.7 80.8 119.0
2180.0 3080.0 3700.0 4610.0 6010.0 2063.6 2934.0 3534.7 4401.4 5697.5 1388.1 1926.2 2401.3 3023.6 4019.3 503.3 796.1 1013.3 1291.6 1894.5 68.8 106.5 162.8 213.2 332.6 4.8 9.1 13.3 19.3 36.0 7.2 15.7 28.3 42.6 92.5 28.0 53.6 89.9 130.6 288.3 11.4 28.8 47.5 74.1 127.8 23.4 42.2 62.8 87.4 135.5 28.0 43.5 62.8 100.3 181.9
1586.0 2220.0 2830.0 3650.0 4956.0 1461.9 2064.0 2691.3 3345.1 4565.6 896.6 1338.0 1739.1 2233.5 3167.0 411.9 630.5 813.2 1053.7 1582.7 63.7 109.3 163.1 219.8 352.8 3.4 6.6 10.4 14.7 25.2 6.7 14.3 25.7 41.2 75.4 30.6 49.0 70.7 95.9 151.5 12.9 30.8 53.5 90.9 177.5 18.5 33.2 49.7 67.8 134.4 24.7 45.1 67.1 99.2 174.8
72 1329.0 1770.0 2102.0 2571.8 3560.0 1170.2 1593.1 1916.9 2328.9 3295.1 685.2 949.5 1177.0 1450.3 2067.5 364.1 537.4 669.9 857.0 1227.1 82.3 132.2 178.2 248.2 377.7 2.1 4.4 6.9 10.4 18.5 6.9 14.3 22.3 33.7 62.2 23.6 43.9 65.9 93.6 164.7 22.8 50.9 77.1 118.4 212.9 13.6 25.4 37.7 54.3 94.4 25.5 46.8 69.5 104.0 187.0
RI PT
1478.5 2187.5 2695.0 3457.5 4745.5 1403.3 2130.3 2607.5 3318.6 4598.5 957.5 1446.6 1848.8 2338.7 3085.2 351.4 542.7 736.4 957.6 1454.5 31.5 63.1 87.7 126.0 225.1 5.4 11.5 17.6 26.0 43.5 3.2 7.6 12.6 19.4 39.8 22.9 43.1 62.5 89.7 164.1 4.0 8.1 12.2 19.7 35.0 7.6 18.6 30.7 43.8 69.2 15.1 20.1 29.7 38.7 53.6
EP
CD4+ TCRγδ+ T cells
5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th
M AN US C
TCRαβ+ T cells
0
D
T cells
Age category (months) 6 14 25
percentile
TE
Supplemental Table 2 (continued) Lymphocytes
ACCEPTED MANUSCRIPT
CD4+ Tnaive
CD4+ Tmem
CD4+ Tcm
EP
TE
CD4+ TemRO
AC C
CD4+ early TemRO
CD4+ intermediate TemRO
CD4+ late TemRO
CD4+ TemRA
CD4+ early TemRA
Age category (months) 6 14 25 5.2 9.0 8.7 11.8 18.3 20.9 22.9 31.7 39.2 38.4 58.0 66.0 71.0 116.5 135.7 1820.5 1430.2 922.3 2706.7 1991.4 1362.5 3366.2 2457.8 1782.7 4101.9 3062.0 2288.7 5553.1 4123.0 3217.7 1519.9 941.0 575.3 2233.6 1469.5 916.6 2849.1 1972.0 1256.0 3460.9 2461.5 1662.9 4841.4 3373.4 2336.6 354.1 331.6 325.1 470.8 526.9 504.6 580.2 669.0 651.4 713.2 834.9 850.9 953.0 1227.9 1274.9 134.9 111.0 86.7 202.7 176.0 129.5 261.9 242.8 173.1 322.1 313.2 223.6 449.3 427.4 349.3 61.2 31.8 76.9 93.0 123.7 152.4 138.3 184.3 205.4 177.5 254.7 260.3 260.8 387.9 387.4 50.8 20.6 54.6 79.0 95.6 116.7 110.3 148.8 154.8 144.8 199.8 197.5 211.9 301.6 276.4 2.7 2.6 9.7 10.2 14.0 17.7 17.3 23.0 26.2 24.7 33.2 36.4 38.5 53.4 64.0 0.0 0.0 1.4 0.0 2.5 3.8 0.0 5.1 7.7 3.8 11.2 17.6 14.3 29.0 43.4 79.9 50.3 58.1 122.0 115.1 144.0 178.0 221.2 245.8 241.2 343.4 415.3 376.8 570.5 812.5 66.7 33.1 40.3 104.3 91.1 114.3 149.9 189.8 213.6 204.6 304.4 356.7 357.8 542.8 782.6
72 13.3 31.0 49.9 83.9 163.1 737.1 988.2 1240.0 1525.3 2134.0 392.9 621.4 821.9 1069.7 1538.0 344.5 461.7 559.2 679.5 878.5 74.2 112.5 149.0 198.9 285.0 118.3 181.5 235.2 295.5 408.5 82.5 127.1 168.0 210.1 292.4 18.2 29.7 43.9 61.5 105.2 1.1 3.0 6.3 12.9 30.1 55.8 102.4 153.6 227.8 371.7 36.3 67.6 110.7 168.4 317.4
RI PT
CD4+ T cells
0 1.5 3.6 5.7 12.5 24.3 1060.7 1552.3 1972.5 2498.9 3266.2 848.7 1321.1 1695.9 2159.7 2898.2 161.0 259.6 355.9 465.1 754.6 39.5 81.4 121.3 198.3 395.6 23.4 51.5 73.8 107.0 172.6 21.0 47.6 65.0 98.9 155.8 0.0 1.8 3.8 7.0 19.6 0.0 0.0 0.0 0.0 3.3 35.9 89.2 126.8 181.1 312.1 27.9 74.2 113.7 162.7 287.3
M AN US C
Vδ2+Vγ9+ T cells
percentile 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th
D
Supplemental Table 2 (continued)
ACCEPTED MANUSCRIPT
CD8+ T cells
CD8+ Tnaive
CD8+ Tmem
EP
AC C
CD8+ TemRO
TE
CD8+ Tcm
CD8+ early TemRO
CD8+ intermediate TemRO
CD8+ late TemRO
CD8+ TemRA
Age category (months) 6 14 25 0.0 0.0 0.0 0.0 0.0 1.4 3.6 2.7 2.7 5.7 4.6 4.7 11.0 9.0 7.8 0.0 0.0 1.7 0.0 3.3 4.4 3.6 6.9 8.8 6.6 15.2 16.0 17.8 44.5 42.0 527.3 447.2 389.3 836.6 776.2 619.4 1093.1 998.8 805.1 1389.5 1277.1 1038.8 2060.9 1831.8 1571.6 380.4 269.6 205.6 591.2 453.9 363.8 835.1 610.3 465.1 1081.6 809.5 640.3 1632.3 1238.7 1033.7 98.1 102.0 71.6 171.4 181.4 152.3 236.3 268.9 224.3 343.6 432.3 331.4 674.6 888.7 550.0 16.7 13.5 9.5 32.1 26.1 18.6 49.2 39.8 31.0 72.6 68.8 56.3 128.0 140.4 109.2 16.7 17.8 14.0 39.3 43.5 36.0 62.6 74.9 67.5 106.4 151.2 108.1 284.8 362.8 244.3 10.7 7.4 6.4 22.7 22.7 19.6 36.5 38.3 30.5 55.4 68.0 48.7 106.8 145.0 88.0 0.9 0.1 0.5 4.7 4.7 3.4 10.2 11.6 8.2 21.5 29.9 18.1 64.9 98.6 56.1 0.0 0.0 0.8 1.2 3.2 3.7 3.1 8.3 9.2 9.4 22.0 24.5 85.0 94.8 73.9 34.4 33.0 14.8 76.8 69.9 53.0 109.4 122.2 102.8 168.9 224.6 180.9 345.1 467.4 367.6
72 0.8 1.8 3.6 6.9 18.2 1.7 4.7 9.4 17.4 42.6 335.8 509.0 649.7 841.6 1197.8 166.1 287.2 383.6 513.2 768.1 122.4 199.0 261.0 365.3 597.5 5.1 9.3 14.4 21.7 43.8 40.5 75.5 114.9 166.3 281.8 19.4 38.1 56.6 79.9 132.0 3.1 8.5 17.1 30.2 64.0 2.2 6.5 14.0 30.4 72.6 44.5 86.0 129.4 194.2 346.9
RI PT
CD4+ late TemRA
0 0.0 0.0 2.7 5.1 13.5 0.0 0.0 1.5 2.9 10.4 316.6 493.3 655.5 886.4 1319.3 182.4 351.1 487.2 695.2 1037.3 57.2 102.8 149.1 214.9 351.2 10.2 27.7 47.7 71.7 147.6 7.9 17.3 29.6 48.4 98.7 5.5 12.2 22.2 32.9 64.3 0.3 2.1 5.8 11.3 27.7 0.0 0.0 0.0 0.0 0.7 18.9 43.1 62.5 96.3 165.1
M AN US C
CD4+ intermediate TemRA
percentile 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th
D
Supplemental Table 2 (continued)
ACCEPTED MANUSCRIPT
B cells
Bnaive
Bmem
AC C
CD27- Bmem
CD27+ Bmem
IgM+ Bmem
Natural effector
IgMonly
72 17.9 33.5 52.1 76.4 139.8 8.9 20.4 33.8 53.4 108.0 4.4 11.9 24.6 52.2 162.6 321.5 480.0 630.0 765.3 1110.0 250.2 389.5 513.6 639.9 933.7 46.0 74.1 98.3 131.2 200.0 20.8 35.8 48.8 65.9 103.8 9.8 17.0 25.2 37.2 65.9 8.7 16.2 25.0 36.8 61.7 6.9 13.0 19.7 30.2 52.1 1.0 2.4 4.1 7.1 15.7
RI PT
Age category (months) 6 14 25 22.3 13.7 5.5 44.5 36.4 28.8 68.3 62.1 54.1 106.0 101.0 89.9 167.0 211.4 169.2 2.6 2.9 1.1 9.8 11.0 6.8 20.8 22.8 17.5 38.9 46.8 34.7 91.3 114.4 79.7 0.0 1.2 1.2 2.8 5.5 5.4 6.3 11.8 12.9 15.6 36.6 33.0 90.4 187.3 151.6 810.0 710.0 520.0 1310.0 1150.0 860.0 1720.0 1490.0 1130.0 2260.0 1950.0 1440.0 3277.5 2590.0 1890.0 733.9 621.0 466.0 1190.7 1001.1 772.7 1609.4 1339.5 1019.3 2104.3 1757.7 1292.0 3030.8 2310.1 1731.0 47.2 59.1 48.8 79.3 98.4 78.3 118.1 142.1 108.2 171.6 191.7 151.5 291.3 305.1 212.6 20.7 27.4 29.4 38.7 52.6 44.2 62.1 72.2 62.8 88.6 98.6 90.5 153.7 157.5 130.0 4.1 6.2 6.3 10.0 18.4 12.0 16.3 28.1 21.1 27.7 42.5 30.9 65.2 75.3 57.1 11.4 12.2 8.9 24.5 26.0 15.5 41.8 41.2 24.7 65.9 68.0 38.1 131.2 133.8 70.2 8.8 9.6 6.9 19.3 20.1 12.3 33.7 33.3 20.4 53.3 55.8 31.3 107.4 113.5 58.3 1.0 0.8 0.9 2.8 3.6 2.4 5.7 6.0 4.0 10.8 10.1 6.5 27.2 25.2 15.6
M AN US C
CD8+ late TemRA
EP
CD8+ intermediate TemRA
0 11.7 29.2 48.0 75.7 123.8 1.4 6.7 11.5 18.8 40.9 0.0 0.0 0.4 1.0 2.1 240.0 467.5 675.0 930.0 1663.5 260.8 425.6 616.3 854.8 1500.9 16.2 31.4 48.0 81.2 168.4 7.7 19.3 29.6 51.7 116.2 0.0 1.2 1.7 3.7 7.0 3.1 6.8 13.0 24.7 67.3 2.7 6.3 12.5 22.3 65.2 0.0 0.0 0.6 1.3 3.6
TE
CD8+ early TemRA
percentile 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th 5th 25th 50th 75th 95th
D
Supplemental Table 2 (continued)
ACCEPTED MANUSCRIPT Age category (months) percentile 0 6 14 25 72 5th 5.8 16.9 16.4 15.0 11.6 25th 11.8 41.3 38.0 30.7 20.4 50th 21.6 70.2 60.8 47.5 28.9 IgA+ Bmem 75th 36.3 117.1 90.9 75.2 40.0 95th 67.9 225.2 151.6 131.3 68.1 5th 4.7 11.3 10.4 10.1 5.0 25th 10.3 31.2 25.3 20.3 10.2 50th 19.0 55.8 43.4 32.4 16.0 CD27-IgA+ 75th 31.9 94.5 67.3 55.6 23.8 95th 62.9 196.7 119.1 101.9 42.6 5th 0.0 3.0 3.3 3.5 4.5 25th 0.9 7.4 9.8 8.1 8.4 50th 1.7 12.0 16.3 13.2 12.2 CD27+IgA+ 75th 3.1 20.1 26.7 21.6 17.3 95th 7.9 42.2 44.2 37.1 27.9 5th 4.2 30.9 33.1 29.7 31.3 25th 8.5 61.0 74.3 61.6 55.7 50th 13.3 87.5 116.8 94.0 77.0 IgG+ Bmem 75th 22.4 143.4 187.6 160.8 102.3 95th 47.9 229.2 318.2 280.1 158.2 5th 3.6 23.3 25.9 23.8 19.2 25th 7.1 48.7 52.6 45.6 36.8 50th 11.2 76.6 87.7 75.4 53.8 CD27-IgG+ 75th 19.2 124.2 151.0 129.2 74.1 95th 42.2 200.3 274.1 224.0 119.3 5th 0.0 3.4 4.4 5.2 7.7 25th 0.8 8.1 14.7 12.6 14.7 50th 1.6 12.3 23.0 19.6 21.1 CD27+IgG+ 75th 3.3 19.2 37.9 33.4 30.1 95th 7.2 35.3 60.8 58.8 51.0 5th 22.0 66.2 60.4 43.1 22.0 25th 38.8 113.7 98.2 74.5 38.4 50th 57.8 157.2 142.8 100.3 55.2 CD21low B cells 75th 91.1 222.4 193.7 134.8 78.1 95th 166.2 351.1 282.9 220.0 124.0 5th 115.0 323.7 290.3 203.9 127.0 25th 220.7 545.5 500.0 351.5 188.8 50th 313.6 757.4 638.0 476.5 251.3 Igλ+ B cells 75th 417.7 975.7 840.0 636.2 317.4 95th 798.4 1486.1 1186.2 847.7 477.2 5th 126.1 456.6 404.8 293.4 179.0 25th 248.8 718.2 635.8 490.5 272.3 50th 352.2 967.0 813.6 634.5 357.0 Igκ+ B cells 75th 493.0 1263.5 1101.9 821.5 439.7 95th 841.4 1857.9 1429.7 1057.4 621.4 The presented reference values are obtained by cross-sectional analyses of all 2,010 data points on the indicated time points
AC C
EP
TE
D
M AN US C
RI PT
Supplemental Table 2 (continued)
ACCEPTED MANUSCRIPT Supplemental Table 3. Overall effect estimates and association with specific age-periods Innate leukocytes Models with 3 knots Determinant
Significance
Granulocytes CD15+ granulocytes Neutrophils Eosinophils Monocytes Classical monocytes Intermediate monocytes
Gender (girl yes/no) Gender (girl yes/no) Low maternal education level Low maternal education level Low maternal education level Seropositivity for CMV at 6yr Siblings >1
** ** ** ** ** ** ****
Determinant
Significance
Antibiotics/Penicillin use in 1st yr
**
NK cells
Effect estimate 359.6 296.4 -63.1 -51.5 -34.3 34.7 -53.2
Standard error 111.6 106.6 24.5 18.1 13.0 12.9 12.1
Effect estimate 6.0
Standard error 2.2
Standard error 120.1 36.1 22.1 20.4 4.6 1.3 1.3 1.7 2.9 3.1 7.2 1.9 37.1 32.4 4.6 5.4 5.0 3.8 3.8 5.3 1.4 21.9 20.3 11.0 10.2
Leukocyte population Non-classical monocytes Lymphocytes Models with 3 knots
M AN US C
Models with 2 knots
Determinant
Significance
T cells TCRαβ+ T cells CD4+ TCRαβ+ T cells
Carrier of H. pylori within 6yrs Gender (girl yes/no) Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Gender (girl yes/no) Low maternal education level Seropositivity for CMV at 6yr Caesarian section Low maternal education level Breastfeeding at 6 months of age Premature rupture of membranes Seropositivity for CMV at 6yr Gender (girl yes/no) Gender (girl yes/no) Gender (girl yes/no) Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Gender (girl yes/no) Seropositivity for EBV at 6y Caesarian section Seropositivity for CMV at 6yr Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Anti-TPO (before 18 weeks of pregnancy) (mU/ml) Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Seropositivity for EBV at 6yr Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Gender (girl yes/no) Seropositivity for CMV at 6yr Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Gender (girl yes/no) Seropositivity for CMV at 6yr
* ** *** * * ** *** ** ** * ** **** ** * ** *** ** ** * * *** *** * **** ****
Effect estimate 303.4 96.9 79.2 51.4 -11.7 3.9 4.9 -4.8 7.9 -6.2 19.5 9.6 107.9 81.3 12.1 20.1 15.5 11.2 9.8 -13.1 5.4 82.3 47.1 104.3 41.2
***
0.02
**** **** **** **** **** **** ** ** **** **** ** *** ****
41.0 31.3 14.1 7.7 9.2 24.8 4.8 -18.1 64.9 8.0 5.5 -13.7 53.3
TCRγδ+ T cells CD4+ TCRγδ+ T cells CD8+ TCRγδ+ T cells Vδ1-Vδ2- TCRαβ- T cells
Vδ1+ T cells CD4+ T cells CD4+ Tnaive CD4+ Tmem CD4+ Tcm CD4+ TemRO
AC C
CD4+ early TemRO CD4+ interm TemRO CD8+ T cells CD8+ Tnaive CD8+ Tmem CD8+ Tcm
EP
TE
Vδ2+ T cells
D
Leukocyte population
CD8+ TCRαβ+ T cells
CD8+ TemRO CD8+ early TemRO CD8+ interm TemRO CD8+ late TemRO CD8+ TemRA CD8+ early TemRA CD8+ interm TemRA CD8+ late TemRA
Significance age-intervals 0-6m 6-14m 14-24m 24-76m ** * **** ns ns ns ns ns ns ns ns ns ns * ** ns * ns ** ns ns ns ns ns **** ns **** ns
RI PT
Leukocyte population
0-14.1m
14.1-70m
70-76m
ns
ns
ns
0-6m
6-14m
14-24m
24-76m
ns ns ns ns ns ns ** ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns * ns
ns ns ns * ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns * ns ns
ns ns * * ns ** ns ns ns ns ns ** ns ns ** ns * ns * ns ns ** ** **** *
* ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns
0.007
****
****
**
****
6.2 5.8 2.6 1.6 1.5 1.9 1.8 6.0 6.5 2.0 1.9 3.6 3.9
ns ns ns ns ns ** ns ** ns ns ns ** **
ns ns ns ns ns ns ** ns ns ns ns * ns
* * * * ** *** ns ns **** ns * ns ****
ns ns ns ns ns ns ns ns ns ns ns ns ns
ACCEPTED MANUSCRIPT Supplemental Table 3 (Lymphocytes; models with 3 knots continued)
Bmem CD27- Bmem CD27+ Bmem IgM+ Bmem Natural effector IgMonly IgA+ Bmem CD27-IgA+ CD27+IgA+ IgG+ Bmem CD27-IgG+ CD27+IgG+ CD21low B cells Igλ+ B cells Igκ+ B cells
Significance
Seropositivity for HSV-1 at 6yr Seropositivity for HSV-1 at 6yr Low maternal education level Seropositivity for EBV at 6yr Low maternal education level Seropositivity for EBV at 6yr Seropositivity for EBV at 6yr Breastfeeding at 6 months of age Carrier of H. pylori within 6yrs Breastfeeding at 6 months of age Carrier of H. pylori within 6yrs Gender (girl yes/no) Carrier of H. pylori within 6yrs Antibiotics/Penicillin use in 1st yr Gender (girl yes/no) Low maternal education level Seropositivity for HSV-1 at 6yr Seropositivity for HSV-1 at 6yr Carrier of H. pylori within 6yrs
** *** * ** * ** * * **** ** **** ** * * ** *** ** ** **
Models with 2 knots
CD4+ TemRA CD4+ early TemRA CD4+ interm TemRA Models with 1 knot Leukocyte population
Significance
Seropositivity for CMV at 6yr Seropositivity for EBV at 6yr Premature rupture of membranes Gender (girl yes/no) Premature rupture of membranes
**** *** ** ** **
D
CD4+ late TemRO
Determinant
TE
Leukocyte population
Determinant
Significance
Gender (girl yes/no) *** Seropositivity for CMV at 6yr **** Seropositivity for EBV at 6yr * Vγ9+ T cells Premature rupture of membranes ** Vδ2+Vγ9+ T cells Premature rupture of membranes *** ns= not significant; *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001
AC C
EP
CD4+ late TemRA
Effect estimate -111.2 -108.8 -8.4 -9.3 -4.7 -6.2 -2.7 -5.8 22.0 -4.9 17.8 1.6 3.1 7.1 2.1 -11.7 -53.0 -59.5 82.2
Standard error 35.8 33.0 3.5 3.4 2.0 2.0 1.1 2.3 4.9 1.9 4.1 0.6 1.2 2.8 0.7 3.2 16.6 19.9 30.9
Effect estimate 6.1 2.7 5.0 -0.7 1.9
Standard error 0.7 0.7 1.8 0.2 0.6
Effect estimate -3.0 7.4 1.8 23.1 24.6
Standard error 0.9 0.9 0.9 7.4 6.6
Significance age-intervals 0-6m 6-14m 14-24m 24-76m ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns * ns * ns ns ns ns ns ns * ns ns * **** ns ns ns ns ns ns * **** ns ns ns ns ns ns ns * ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns **
RI PT
B cells Bnaive
Determinant
M AN US C
Leukocyte population
0-14.1m
14.1-70m
70-76m
ns ns ns ns ns
ns ns **** ns ns
* ns ns ns ns
0-25.5m
25.5-76m
ns **** ns * *
ns * ns ns ns
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
A
EP
TE
D
Supplemental Figure 1
all events
granulocytes
granulocytes
monocytes interm
cells/μl
40000 35000 30000 25000 20000 15000 10000 5000 0
classical
12
24
36
48
3500 3000 2500 2000 1500 1000 500 0
60
72
84
96
cells/μl
35000 30000 25000 20000 15000 10000 5000 0
0
12
24
30000
48
60
72
84
10000 5000 0 36
48
60
3000
72
84
0
96
12
24
900 800 700 600 500 400 300 200 100 0
2000 1500 1000 500 0 0
12
24
36
48
60
36
48
60
72
84
age (months)
data points
96
72
84
0
96
12
24
36
48
60
12
450 400 350 300 250 200 150 100 50 0
72
84
96
60
24
36
48
60
72
84
96
72
84
96
non-classical monocytes
0
12
24
36
48
60
age (months)
age (months)
linked data points
48
age (months)
intermediate monocytes
0
36
eosinophils
age (months)
classical monocytes
2500
24
9000 8000 7000 6000 5000 4000 3000 2000 1000 0
15000
24
12
age (months)
20000
12
0
96
neutrophils
25000
age (months)
cells/μl
36
CD3
NK cells
age (months)
CD15+ granulocytes
0
D
3500 3000 2500 2000 1500 1000 500 0
monocytes
age (months)
C
CD14
CD45
CD45
granulocytes
0
eosinophils
lymphocytes
CD45
B
non-classical
CD16
CD16
AC C
CD15
SSC
monocytes
NK cells
CD16/CD56
neutrophils
CD15+ granulocytes
granulocytes
lymphocytes
model
90% CI
72
84
96
A T cells
EP
TCRγδ
CD4+ TCRαβ+ T cells
AC C
CD16/CD56
CD3
TCRγδ+ T cells
CD4+ TCRγδ+ T cells
Vδ1+ T cells
Vδ2+Vγ9+ T cells
Vγ9+ T cells
Vδ1-Vδ2-
TCRαβ
B
CD4
CD4
Vδ1
Vγ9
C
12000
12000
T cells
10000 8000
cells/μl
cells/μl
TCRαβ+ T cells
TCRγδ+ T cells Vδ2+ T cells
CD8+ TCRγδ+ T cells
CD8
CD8+ TCRαβ+ T cells
TCRγδ+ T cells
T cells
TCRγδ+ T cells
TCRαβ+ T cells
CD8
lymphocytes
Vδ2
Supplemental Figure 2
Vδ2
TE
D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
6000 4000 2000 0
0
12
24
36
48
60
72
84
96
5000
8000
4000
6000
3000
4000
2000
2000 0
1000 0
0
12
24
cells/μl cells/μl
1400 1200 1000 800 600 400 200 0 400 350 300 250 200 150 100 50 0
36
48
60
72
84
96
250
TCRγδ+ T cells
CD8+ TCRγδ+ T cells
200
100 50 12
24
36
48
60
72
84
96
Vδ1+ T cells
0
12
24
36
48
0
60
72
84
96
0 900 800 700 600 500 400 300 200 100 0
0
12
24
36
48
60
72
84
96
Vδ2+ T cells
0
12
24
36
48
60
72
84
96
140 120 100 80 60 40 20 0
0 900 800 700 600 500 400 300 200 100 0 0
age (months)
age (months)
data points
12
24
36
48
60
72
84
96
9000 8000 7000 6000 5000 4000 3000 2000 1000 0
CD4+ TCRαβ+ T cells
0
12
24
linked data points
1400 1200 1000 800 600 400 200 0 0 12 24 36 48 60 72 84 96 800 Vγ9+ T cells 700 600 500 400 300 200 100 0 0 12 24 36 48 60 72 84 96
48
60
72
84
96
Vδ1-Vδ2- TCRαβ- T cells
CD4+ TCRγδ+ T cells
12
24
36
48
60
90% CI
72
84
96
Vδ2+Vγ9+ T cells
12
24
36
48
60
age (months)
age (months)
model
36
age (months)
age (months)
150
0
CD8+ TCRαβ+ T cells
age (months)
age (months)
D
6000
TCRαβ+ T cells
10000
72
84
96
RI PT
ACCEPTED MANUSCRIPT
CD4+ T cells Tcm
CD4+ T cells
CD4+ Tmem
TemRA
CD3
B
CD4+
12
24
36
48
60
72
84
cells/μl
9000 8000 7000 6000 5000 4000 3000 2000 1000 0 0 96
early
interm
late
early
late
CD27
CCR7
AC C
9000 8000 7000 6000 5000 4000 3000 2000 1000 0 0
interm
CD4+ TemRA
CD4+ Tnaive
EP
CD8
CD45RO
TemRO
CD4+ TemRO
CD28
TE
T-cells
CD28
A
D
M AN U
SC
Supplemental Figure 3
CD27 2500
CD4+ Tnaive
CD4+ Tmem
2000 1500 1000 500 12
24
36
48
60
72
84
96
0 0
12 24 36 48 60 72 84 96
age (months)
age (months)
age (months)
cells/μl
C
1600 1400 1200 1000 800 600 400 200 0 0
CD4+ Tcm
12
24
36
48
60
72
84
96
1200 1000 800 600 400 200 0 0
1600 CD4+TemRO 1400 1200 1000 800 600 400 200 0 0 12 24 36 48 60 72 84 96
age (months)
CD4+ TemRA
12 24 36 48 60 72 84 96
age (months)
age (months)
cells/μl
D 700 600 500 400 300 200 100 0 0
CD4+ TemRO interm
200 150 100 50
12
24
36
48
60
72
84
96
0 0
12
24
1600 1400 1200 1000 800 600 400 200 0 0
CD4+ TemRA early
12
24
36
48
60
72
36
48
60
72
84
300 250 200 150 100 50 0 0 96
CD4+ TemRO late
12
24
age (months)
age (months)
E
cells/μl
250
CD4+ TemRO early
84
96
age (months) data points
70 60 50 40 30 20 10 0 0
CD4+ TemRA interm
12
24
36
48
60
72
84
200 180 160 140 120 100 80 60 40 20 0 0 96
48
60
72
84
96
84
96
CD4+ TemRA late
12
24
36
48
60
72
age (months)
age (months) linked data points
36
age (months)
model
90% CI
SC
RI PT
ACCEPTED MANUSCRIPT
T-cells
CD8+ Tmem
0
12
24
36
48
60
72
84
96
cells/µl
600 500 400 300 200 100 0 0
CD8+Tcm
2500
CD8+Tnaive
1000 500
00
12
24
36
48
60
72
84
96
00
age (months) 2500
CD8+TemRO
2000
1000 500 60 72 84 96
CD8+Tmem
2000
1500
12 24 36 48
CD27
1500
age (months)
C
late
CD27
3000 2500 2000 1500 1000 500
CD8+
interm
late
EP
6000 5000 4000 3000 2000 1000 0
interm
CCR7
AC C
cells/µl
B
early
early
CD8+ Tnaive
TemRA
CD8+ TemRA
CD28
CD45RO
CD8
Tcm
TE
TemRO
CD8+ T cells
CD3
CD8+ TemRO
CD8+ T cells
CD28
A
D
M AN U
Supplemental Figure 4
00
12 24 36 48 60 72 84 96
age (months) 1200 1000 800 600 400 200 0
CD8+TemRA
0
12 24 36 48 60 72 84 96
age (months)
age (months)
age (months)
12 24 36 48 60 72 84 96
cells/µl
D 1400 1200 1000 800 600 400 200 0
CD8+TemRO early
0
12 24 36 48 60 72 84 96
400 350 300 250 200 150 100 50 00
CD8+TemRO interm
12 24 36 48 60 72 84 96
450 400 350 300 250 200 150 100 50 00
12 24 36 48 60 72 84 96
age (months)
age (months)
age (months)
CD8+TemRO late
cells/µl
E 500 450 400 350 300 250 200 150 100 50 0
CD8+TemRA early
0
12 24 36 48 60 72 84 96
age (months) data points
400 350 300 250 200 150 100 50 0
CD8+TemRA interm
0
12 24 36 48 60 72 84 96
1000 900 800 700 600 500 400 300 200 100 0 0
age (months) linked data points
CD8+TemRA late
12 24 36 48 60 72 84 96
age (months) model
90% CI
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
12 24 36 48 60 72 84 96
age (months)
0
12 24 36 48 60 72 84 96
450 400 350 300 250 200 150 100 50 0 0
age (months)
cells/μl
cells/μl
H
700 600 500 400 300 200 100 0 0 140 120 100 80 60 40 20 0 0
400 350 300 250 200 150 100 50 0 0 12 24 36 48 60 72 84 96 120 IgMonly 100 80 60 40 20 0 0 12 24 36 48 60 72 84 96
Natural effector
age (months)
350 300 250 200 150 100 50 0 0 12 24 36 48 60 72 84 96
age (months)
IgM
IgG+ Bmem
IgA+ Bmem
IgMonly
CD27
800 700 600 500 400 300 200 100 0 0 12 24 36 48 60 72 84 96
D cells/μl 12 24 36 48 60 72 84 96
CD27+ Bmem
12 24 36 48 60 72 84 96
3000 2500 2000 1500 1000 500 0 0
CD27
700 600 500 400 300 200 100 0 0
12 24 36 48 60 72 84 96
450 400 350 300 250 200 150 100 50 0 0 12 24 36 48 60 72 84 96
800 700 600 500 400 300 200 100 0 0 12 24 36 48 60 72 84 96
IgG+ Bmem
12 24 36 48 60 72 84 96
age (months)
age (months)
Cluster 2 Natural effector CD27-IgA+
2 1
CD27+IgG+
12 24 36 48 60 72 84 96
3
CD27-IgG+
36 48 60 72 84 96
Igλ+ B cells
age (months)
IgA+ Bmem
age (months)
I
00
age (months)
IgM+ Bmem
age (months)
IgM
2500 2000 1500 1000 500
Igκ+ B cells
age (months)
age (months)
G
CD27
Bmem
Bnaive
600 500 400 300 200 100 0 0 12 24 12 24 36 48 60 72 84 96 180 CD27+IgA+ 160 140 120 100 80 60 40 20 0 0 12 24 12 24 36 48 60 72 84 96
CD27-IgA+
natural effector
CD27+
CD27
age (months)
CD27+ Bmem CD27+
CD27+
IgM+ Bmem
CD27- Bmem
B cells CD27-
cells/μl
CD21low B cells
5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0 0
B cells CD27-
IgD
EP linked data points model 90% CI
F
800 700 600 500 400 300 200 100 0
C
data points
cells/μl
cells/μl
E
B cells
B cells
Bmem
CD27-
CD21
cells/μl
6000 5000 4000 3000 2000 1000 0 0
Igĸ
CD21low B cells
AC C
cells/μl
B
CD38
Igλ
SSC
Igĸ+ B cells
B cells
CD19
B cells
Bnaive
IgD
B cells
Igλ+ B cells
IgA
B cells
all events
Z-score
A
IgG
TE
D
Supplemental Figure 5
0
Cluster 3 IgMonly CD27-IgG+ CD27+IgA+
-1 -2 -3
Cluster 4 CD27+IgG+
-4 36 48 60 72 84 96
age (months)
-5
0
6
12
18
24
30
36
42
age (months)
48
54
60
66
72
RI PT
ACCEPTED MANUSCRIPT
2
cl4
cl3
3
cl4
cl3
4
4 3 9 6 8 25 52 55 47 43 19 7 37 23 35 40 39 53 56 58 31 41 45 28 44 32 21 33 27 29 59
10
20
30
40 Age (months)
50
6
3
Z-score
2
2
1 0 -1
Z-score
-2 3 2 1 0 -1 -2 -3 -4 -5 2 1
0
6
EP
4 3 9 6
0
12 18 24 30 36 42 48 54 60 66 72
0
1
2
Eosinophils Neutrophils NK cells Classical monocytes
8 25 52 55 47 43 19 7 37 23 35
Non-classical monocytes CD4+ Tcm Natural effector B cells CD27-IgA+ Bnaive CD8+ TemRA early Vδ1+ T cells Intermediate monocytes CD8+ Tcm CD4+ Tnaive CD8+ Tnaive
40 39 53 56 58 31
CD8+ TemRO interm CD8+ TemRO early IgMonly CD27+IgA+ CD27-IgG+ CD4+ TemRA early
41 45 28 44 32 21 33 27 29 59
CD8+ TemRO late CD8+ TemRA late CD4+ TemRO interm CD8+ TemRA interm CD4+ TemRA interm Vδ2+Vγ9+ T cells CD4+ TemRA late CD4+ TemRO early CD4+ TemRO late CD27+IgG+
cluster 1
3
0
6
12 18 24 30 36 42 48 54 60 66 72
4
0 Z-score
-1
12 18 24 30 36 42 48 54 60 66 72
-1 -2 -3 -4 -5
70
TE
1
AC C
Z-score
-2
6 5 4 3 2 1 0 -1 -2
60
z-score
D
0
M AN U
cl2
cl1
1
SC
Supplemental Figure 6
0
6
12 18 24 30 36 42 48 54 60 66 72 Age (months)
cluster 3
ACCEPTED MANUSCRIPT SUPPLEMENTAL MATERIALS AND METHODS Study subjects Included in our study are 1,182 two-generation Dutch children. Peripheral blood was
RI PT
obtained from 220 children at birth, from 376 children at the age of 5-9 months, from 241 children at the age of 13-17 months, from 257 children at the age of 22-30 months and from 916 children at the age of 61-95 months. Of 237 children, detailed immunophenotyping data was available at 3 or more follow-up time points, of 90 children at 4 or more time points. The
M AN US C
complete follow-up at 5 time points was available for 12 children.
Sample preparation
Approximately 4ml of blood was collected in EDTA-anticoagulant tubes by venous puncture. Blood was kept at room temperature and processed within 24 hours after sampling. 50µl Full blood was mixed with BD Trucount beads (BD Biosciences; San Jose, CA) and stained for CD3, CD16, CD56, CD19 and CD45 in a diagnostic lyse-no-wash approach to obtain absolute cell counts. Subsequently, red blood cells were lysed by adding NH4Cl. Absolute
D
numbers of leukocytes, NK cells, T cells and B cells were obtained without a wash step on a BD FACSCalibur (BD Biosciences; San Jose, CA).
TE
2ml Full blood was subjected to bulk lysis of red blood cells by 10 min incubation with NH4Cl. Following wash steps, the remaining white blood cells were divided over 10
EP
tubes for 6-color flowcytometric analyses with the antibodies listed in Supplemental Table 1. Following 10 min incubation at room temperature, cells were washed and immunophenotypic
AC C
measurements were performed on a 3-laser BD LSRII (BD Biosciences; San Jose, CA). The 488nm laser was used to excite the FITC (530/30nm filter), PE (575/26nm), PerCP-Cy5.5 (695/40nm) and PE-Cy7 (780/60nm) fluorophores, and the 633nm laser for APC (660/20nm) and APC-Cy7 (780/60nm). PMT voltages were set using 8-peak rainbow beads based on target values for the 7th peak as established within the EuroFlow consortium.27 Absolute cells per µl blood of leukocyte subpopulations were recalculated using the total leukocyte, NKcell, T-cell and B-cell numbers obtained from the Trucount analysis.
Statistical modeling Our dataset included 62 leukocyte populations and 26 determinants for 1,182 children, with for each leukocyte population (defined as population “i”), a total of 2,010 data points.
Linear mixed effect modeling of leukocyte kinetics 1
ACCEPTED MANUSCRIPT For each leukocyte population “i” in our dataset, the following 4 linear mixed effect models were tested:
lmer(dataset[,i] ~ ns(Age,1)+(1|Id),REML=FALSE,data=data)
lmR_1knots<-
RI PT
lmR_0knots<-
M AN US C
lmer(dataset[,i] ~ ns(Age,2, knots = c(25.5))+(1|Id),REML=FALSE,data=dataset)
lmR_2knots<-
lmer(dataset[,i] ~ ns(Age,3, knots = c(14.1 , 70))+(1|Id),REML=FALSE,data=dataset)
lmR_3knots<-
lmer(dataset[,i] ~ ns(Age,4, knots = c(6, 14, 24))+(1|Id),REML=FALSE,data=dataset)
# ns(Age,...) defines the natural spline of the age of the children, with indicated knots;
TE
longitudinal measurements.
D
# (1|Id) defines the child’s Id-number as a random effect in the model, to include
EP
The optimal model was selected by likelyhood ratio test:
anov[1]<-ifelse(anova(lmR_0knots,lmR_1knots)[[8]][2]<0.05,2,0)
AC C
anov[2]<-ifelse(anova(lmR_1knots,lmR_2knots)[[8]][2]<0.05,3,0) anov[3]<-ifelse(anova(lmR_2knots,lmR_3knots)[[8]][2]<0.05,4,0) selected.lmR<-max(anov)
Univariate analyses of the effect of nongenetic determinants on leukocyte kinetics Each determinant “j” in our dataset was included independently as a fixed effect to the optimal linear mixed effect model of individual leukocyte populations (below an example for a leukocyte population modelled by model lmR_3knots) lmer(dataset[,i] ~ ns(Age,4, knots = c(6,14,24))+determinant[,j]+(1|Id),REML=FALSE,data=dataset)
Determinants with an effect of p<0.0125 were defined as significant. 2
ACCEPTED MANUSCRIPT Multivariable analyses of the effect of nongenetic determinants on leukocyte kinetics All determinants with a significant univariate effect on a leukocyte population (up to three determinants per leukocyte population), were combined and included as fixed effects to the
RI PT
optimal linear mixed effect model of individual leukocyte populations to correct for confounding effects (below an example for a leukocyte population modelled by model lmR_3knots)
M AN US C
lmer(dataset[,i] ~ ns(Age,4, knots = c(6,14,24))+significant determinant 1+ significant determinant 2+ etc +(1|Id),REML=FALSE,data=dataset)
Determinants with an effect of p<0.05 were defined as significant after multivariable correction.
Define whether a determinant has an age-associated affect on a leukocyte population For determinants that still affected leukocyte kinetics after multivariable correction, we
D
subsequently defined whether this effect was significantly stronger within a selective ageperiod (either between the age of 0months – knot1, knot1 – knot2 , knot2 – knot3 or knot3 –
TE
76 months). To test this, the models for multivariable analyses were adjusted, testing
EP
individual determinants in relation to the age in the children:
e.g. Test the age-associated effect of determinant 1: lmer(dataset[,i] ~ ns(Age,4, knots = c(6,14,24))* significant determinant 1 + significant
AC C
determinant 2+ etc +(1|Id),REML=FALSE,data=dataset)
# “*” indicates a test of the natural spline of the age of the children in relation to determinant 1 # the other determinants significantly affecting this leukocyte population were still included as fixed effects to correct for potential confounding effects
3
ACCEPTED MANUSCRIPT
M AN US C
RI PT
SUPPLEMENTAL FIGURES: FIGURE LEGENDS Supplemental Figure 1. Dynamics of innate leukocyte subsets in children between birth and 6 years of age. A) Gating strategy used to identify innate leukocyte subsets. Plots depict data of a representative 61-95 months-old child. B-D) Absolute numbers and modelled dynamics of granulocytes, monocytes and NK cells (B), CD15+, neutrophilic and eosinophilic granulocytes (C), and classical, intermediate and non-classical monocyte subsets (D), from birth until 6 years of age. Linear mixed effect models were generated for each population (solid black line) and represented with the 90% confidence interval (CI) of the model (dashed black lines). For clarity of the graphs, gray lines connect only consecutive time points within one individual; i.e. 0-6m, 6-14m, 14-25m or 25-76m.
D
Supplemental Figure 2. Dynamics of TCRαβ+ T-cell and TCRγδ+ T-cell subsets in children between birth and 6 years of age. A) Gating strategy used to identify TCRαβ+ T-cell and TCRγδ+ T-cell subsets. Plots depict data of a representative 61-95 months-old child. B-D) Absolute numbers and modelled dynamics of T cells (B), TCRαβ+ T-cell populations (C), and TCRγδ+ T-cell populations (D), from birth until 6 years of age. Linear mixed effect models were generated for each population (solid black line) and represented with the 90% confidence interval (CI) of the model (dashed black lines). For clarity of the graphs, gray lines connect only consecutive time points within one individual; i.e. 0-6m, 6-14m, 14-25m or 25-76m.
AC C
EP
TE
Supplemental Figure 3. Dynamics of CD4+ T-cell subsets in children between birth and 6 years of age. A) Gating strategy used to identify CD4+ T-cell subsets. Plots depict data of a representative 61-95 months-old child. B-E) Absolute numbers and modelled dynamics of total, naive and memory CD4+ T cells (B); CD4+ central memory T cells (Tcm), CD45RO+ effector memory (TemRO) and CD45RO- effector memory (TemRA) T-cell subsets (C); early, intermediate or late differentiated TemRO subsets (D); or early, intermediate or late differentiated TemRA subsets (E), between birth and 6 years of age. Linear mixed effect models were generated for each population (solid black line) and represented with the 90% confidence interval (CI) of the model (dashed black lines). For clarity of the graphs, gray lines connect only consecutive time points within one individual; i.e. 0-6m, 6-14m, 14-25m or 25-76m. Supplemental Figure 4. Dynamics of CD8+ T-cell subsets in children between birth and 6 years of age. A) Gating strategy used to identify CD8+ T-cell subsets. Plots depict data of a representative 61-95 months-old child. B-E) Absolute numbers and modelled dynamics of total, naive and memory CD8+ T cells (B); CD8+ central memory T cells (Tcm), CD45RO+ effector memory (TemRO) and CD45RO- effector memory (TemRA) T cell subsets (C); early, intermediate or late differentiated TemRO subsets (D); or early, intermediate or late differentiated TemRA subsets (E), between birth and 6 years of age. Linear mixed effect models were generated for each population (solid black line) and represented with the 90% confidence interval (CI) of the model (dashed black lines). For clarity of the graphs, gray lines connect only consecutive time points within one individual; i.e. 0-6m, 6-14m, 14-25m or 25-76m. Supplemental Figure 5. Dynamics of B-cell subsets in children between birth and 6 years of age. 4
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
RI PT
A) Gating strategy used to identify B-cell subsets. Plots depict data of a representative 61-95 months-old child. B-I)Absolute numbers and modelled dynamics of total B cells (B); naive and total memory B cells (C); total Igκ+ and Igλ+ B-cells populations (D); and CD21low B cells (E); CD27- and CD27+ memory B cells (F); total IgM+, IgA+ or IgG+ memory B cells (G); IgM+, IgA+ or IgG+ memory B-cell subsets (H) between birth and 6 years of age. Linear mixed effect models were generated for each population (solid black line) and represented with the 90% confidence interval (CI) of the model (dashed black lines). For clarity of the graphs, gray lines connect only consecutive time points within one individual; i.e. 0-6m, 614m, 14-25m or 25-76m. I) Overlay of the normalized (zero mean; unit SD) linear mixed effect models of the memory B-cell populations as in panel H.
5
ACCEPTED MANUSCRIPT
AC C
EP
TE
D
M AN US C
RI PT
Supplemental Figure 6. Hierarchical clustering of the dynamics of 31 non-overlapping leukocyte subsets in early childhood. A) Ward’s hierarchical clustering was performed as in Figure 2, including only the 31 nonoverlapping leukocyte subsets. Indicated in gray squares in front of the heat map is the cluster that each population was assigned to in the analysis of the total 62 leukocyte subsets in Figure 2. B) Average patterns +/- 1 standard deviation of the subsets in each of the 4 major clusters. Indicated in black squares are 3 subsets that were assigned to different clusters than in Figure 2.
6